Support the modernization, expansion and adaptation to the future of mobility of the traffic lights systems with the advent of 5G technology in São Paulo Task C - Diagnosis Selection #1265335 TASK C: DIAGNOSIS OF CURRENT TECHNOLOGY AND THE CURRENT TRAFFIC AND TELECOM NETWORKS FOR THE CASE STUDY IN SÃO PAULO SUPPORT THE MODERNIZATION, EXPANSION AND ADAPTATION TO THE FUTURE OF MOBILITY OF THE TRAFFIC LIGHTS SYSTEMS WITH THE ADVENT OF 5G TECHNOLOGY IN SÃO PAULO CUSTOMER THE WORLD BANK GROUP FUTURE MOBILITY FOR SÃO PAULO CONSORTIUM June, 2022 C - Diagnosis DOCUMENT CONTENT MANAGEMENT Status of the Date of Processed by Document Version Comments / Modifications update Content 09/30/2021 FMfSP First delivery 1 10/25/2021 FMfSP Review 2 Recommendations from WB Inclusion of maps in Main Document and 10/28/2021 FMfSP Review 3 individual Acronyms List/Glossary Inclusion of information for Infovia and 5G 01/15/2022 FMfSP Review 4 Auction, and new iteration of Dimensions C - Diagnosis TABLE OF CONTENTS 1 INTRODUCTION 14 2 GENERAL DATA 17 2.1 CITY OF SÃO PAULO 17 2.2 INSERTION IN THE METROPOLITAN REGION 17 2.3 ORGANIZATION OF PUBLIC AND PRIVATE SECTOR 19 2.3.1 São Paulo State Government 19 2.3.2 Metrô – São Paulo Subway Company 19 2.3.3 CPTM – São Paulo Metropolitan Train Company 19 2.3.4 PMSP – São Paulo City Hall 20 2.3.5 SMT – Municipal Secretary of Mobility and Traffic 20 2.3.6 CET – Traffic Engineering Company 20 2.3.7 SPTrans – São Paulo Transport 21 2.3.8 SMIT – Municipal Secretary of Innovation and Technology 21 2.3.9 Anatel – National Telecommunications Agency 22 2.3.10 MNOs – Mobile Network Operators 22 3 TELECOMMUNICATION IN BRAZIL 24 3.1 5G AUCTION IN BRAZIL 26 3.2 NEUTRAL HOST 30 3.2.1 What is a Neutral Host Network? 30 3.2.2 Related Regulation in Brazil 31 3.3 TELECOMMUNICATIONS IN SÃO PAULO 32 4 MOBILITY GENERAL DATA 37 4.1 MODAL SPLIT 37 4.2 VEHICLE FLEET 38 4.3 ROAD CRASHES 42 5 TRAFFIC LIGHTS DATA COLLECTION 45 5.1 TRAFFIC LIGHTS PPMI 45 5.2 CONSULTANCY SERVICES FOR ALTERNATIVES ANALYSIS AND OTHER STUDIES FOR TRAFFIC LIGHTS SYSTEMS IN SÃO PAULO (IDOM STUDY) 47 5.3 MODERNIZATION AND MAINTENANCE TENDER 48 5.4 SMART CITY PMI 51 6 TRAFFIC LIGHTS CONTRACTS 53 7 TRAFFIC LIGHTS CHARACTERIZATION 58 7.1 INPUTS FOR TRAFFIC LIGHTS CHARACTERIZATION 58 7.1.1 Road and Traffic Light System Diagnosis 58 7.1.1.1 Road Hierarchy 58 3 of 134 C - Diagnosis 7.1.1.2 Traffic Light Network 61 7.1.1.3 Controllers 71 7.1.1.4 Detectors 79 7.1.2 Public Transport Diagnosis 80 7.1.2.1 Subway 80 7.1.2.2 Metropolitan Train 84 7.1.2.3 City Bus 85 7.1.2.4 Individual public transport 91 7.1.3 Communication Diagnosis 92 7.1.3.1 CET’s Communication Infrastructure 94 7.1.3.2 Connection Type 96 7.1.4 Micro Mobility Diagnosis 97 7.1.4.1 Bikes 97 7.1.4.2 Pedestrians 103 7.1.4.3 Bike and Scooter Sharing 108 7.1.5 Mobility and Territory Additional Inputs 109 7.1.5.1 Quantity of Jobs 109 7.1.5.2 Land Uses 110 7.1.5.3 OD Survey 111 7.1.5.4 Population Density 112 7.2 METHODOLOGY FOR TRAFFIC LIGHTS CHARACTERIZATION 113 7.2.1 Dimension 1: Current available infrastructure 113 7.2.1.1 Evaluation indicators used: 113 7.2.1.2 Methodology description 114 7.2.1.3 Results – Dimension 1 116 7.2.2 Dimension 2: Infrastructure usage trends 120 7.2.2.1 Evaluation indicators used: 120 7.2.2.2 Methodology description 121 7.2.2.3 Results – Dimension 2 121 7.2.3 Dimension 3: Communication and connection to central traffic control 125 7.2.3.1 Evaluation indicators used: 125 7.2.3.2 Methodology description 125 7.2.3.3 Results – Dimension 3 125 8 GENERAL CONCLUSIONS 130 9 BIBLIOGRAPHY 133 4 of 134 C - Diagnosis LIST OF TABLES Table 1. 5G implementation plan in Brazil 28 Table 2. 5G Auction results and permit values 29 Table 3. Pros and cons of a Neutral Host Scenario 31 Table 4. Mobile network coverage in São Paulo 32 Table 5. Quantity of BTSs (Base Transceiver Stations) per district – top 5 and bottom 5 33 Table 6. Quantity of trips originated in/attracted to São Paulo city by main mode of travel 37 Table 7. São Paulo’s vehicle fleet, quantity per type 39 Table 8. Percentage of families that own a private car, divided per family income 40 Table 9. Number of vehicles divided by year of model launch, city of São Paulo 41 Table 10. Relation between crashes and traffic light intersections (city of São Paulo) 44 Table 11. Types of expression of interest (PMI and PPMI) 45 Table 12. Comparison of bidding processes for the last three traffic light maintenance contracts 53 Table 13. Traffic lights maintenance contracts in force 56 Table 14. Pricing of services for traffic light maintenance, contracts nº 062/17 (lot 01), nº 063/17 (lot 02) and nº 064/17 (lot 03) 56 Table 15. Quantity of signalized intersections by road hierarchy 59 Table 16. Identification of Traffic Engineering Management units (GETs) and their respective Traffic Engineering Departments (DETs) 62 Table 17. Identification and location of Traffic Light Control Departments (DCSs) 63 Table 18. Quantity of regular and flashing yellow traffic lights in each GET 67 Table 19. Type of failures and priority classification 69 Table 20. Quantity of failures by priority (August 2017 – August 2018) 70 Table 21. Quantity of controllers in each GET 72 Table 22. Quantity of centralized controllers in each DCS 72 Table 23. Quantity of controllers with push buttons for pedestrians 72 Table 24. Quantity of controllers that are centralizable and non-centralized in each GET 73 Table 25. Quantity of controllers by supplier in each GET 75 Table 26. Quantity of controllers by supplier in each GET 76 Table 27. Quantity of detectors supply in the Modernization and Maintenance Tender (2019) 80 Table 28. Passenger entry by line in March 2021 83 Table 29. Quantity of signalized intersections in bus corridors (existing and planned) 89 Table 30. Quantity of signalized intersections in exclusive lanes for buses 90 Table 31. Quantity of signalized intersections considering possibility of centralization and public transport corridors. 90 Table 32. Infovia characteristics 93 Table 33. Communication infrastructure length by connection type 94 Table 34. Quantity of controllers by communication infrastructure 96 5 of 134 C - Diagnosis Table 35. Quantity of controllers by connection type 97 Table 36. Quantity of signalized intersections with signal heads for cyclists 98 Table 37. Quantity of signalized intersections considering the cycling network 101 Table 38. Cyclists per family income in São Paulo 101 Table 39. Quantity of signalized intersections with signal heads for pedestrians 106 Table 40. Quantity of signalized intersections considering pedestrian signal heads and push buttons 107 Table 41. Quantity of signalized intersections in PEC routes 107 Table 42. Quantity of signalized intersections considering the pedestrian network 108 Table 43. Quantity of jobs per district – top 5 and bottom 5 110 Table 44. Job index per district – top 5 and bottom 5 110 Table 45. Groups for use of land 111 Table 46. Groups for daily trips by main mode of travel 111 Table 47. Quantity of intrazonal trips by foot – top 5 and bottom 5 112 Table 48. Population density by district – top 5 and bottom 5 112 Table 49. First cluster Typology (XL-intersections) 114 Table 50. Second Cluster Typology (L-intersections) 115 Table 51. Third Cluster Typology (M-intersections) 115 Table 52. Fourth Cluster Typology (S-intersections) 116 Table 53. Fifth Cluster Typology (XS-intersections) 116 Table 54. Results of dimension 1 116 Table 55. Results of dimension 2 121 Table 56. Results of dimension 3 126 LIST OF FIGURES Figure 1. Schematic chronogram and overview of reports 15 Figure 2. Diagnosis Flowchart 16 Figure 3. Administrative Division Map 18 Figure 4. Market share of mobile phone services in the city of São Paulo 22 Figure 5. Number of mobile phone accesses in the city of São Paulo 23 Figure 6. Quantity of internet wide band accesses by provider in Brazil 25 Figure 7. Quantity of mobile phone accesses by provider in Brazil 25 Figure 8. Quantity of cable TV accesses by provider in Brazil 26 Figure 9. Quantity of fixed telephony accesses by provider in Brazil 26 Figure 10. BTS and 4G Coverage Map 35 Figure 11. Modal split in São Paulo city 38 Figure 12. São Paulo’s vehicle fleet per type 39 6 of 134 C - Diagnosis Figure 13. Quantity of vehicles divided by year of model launch, city of São Paulo 42 Figure 14. Annual evolution of fatal crashes by type 43 Figure 15. Vehicles involved in fatal crashes 44 Figure 16. Chronology of data sources 45 Figure 17. Modernization and Maintenance Tender Map 50 Figure 18. Road Hierarchy Map 60 Figure 19. Proportion of signalized intersections by road hierarchy 61 Figure 20. Management Areas Map 65 Figure 21. Proportion of regular and flashing yellow traffic lights 67 Figure 22. Traffic Light System Map 68 Figure 23. Quantity of failures by priority (August 2017 – August 2018) 71 Figure 24. Proportion of controllers that are centralizable and non-centralized 73 Figure 25. Controllers Map 74 Figure 26. Controller Supplier Map 77 Figure 27. Proportion of controllers by supplier 78 Figure 28. Proportion of controllers by age 78 Figure 29. Diagram of São Paulo’s metropolitan transport 81 Figure 30. Subway and Train Network Map 82 Figure 31. Flow of passengers transported by subway (January 2020 – March 2021) 83 Figure 32. City Bus Network Map – Bus corridors and Bus Exclusive Lanes 86 Figure 33. City Bus Network Map – Bus Lines 87 Figure 34. Flow of passengers transported by the city bus fleet (January 2020 – March 2021) 88 Figure 35. Proportion of signalized intersections in bus corridors (existing and planned) 89 Figure 36. Proportion of signalized intersections in exclusive lanes for buses 90 Figure 37. Duct Network Map 95 Figure 38. Proportion of controllers by communication infrastructure 96 Figure 39. Proportion of controllers by connection type 97 Figure 40. Cycling Network Map 99 Figure 41. Traffic Light System Map – Cyclists 100 Figure 42. Proportion of signalized intersections with signal heads for cyclists 101 Figure 43. Cyclist counting (both directions) on bicycle lanes located at Brigadeiro Faria Lima Avenue, Vergueiro Street, and Dr. Gastão Vidigal Avenue 102 Figure 44. Traffic Light System Map – Pedestrians 105 Figure 45. Proportion of signalized intersections with signal heads for pedestrians 106 Figure 46. Proportion of signalized intersections in PEC routes 107 Figure 47. Dimension 1 Map 118 Figure 48. Districts summary – Dimension 1 Map 119 7 of 134 C - Diagnosis Figure 49. Proportion of intersections in dimension 1 120 Figure 50. Proportion of districts in dimension 2 122 Figure 51. Dimension 2 Map 123 Figure 52. Districts Summary – Dimension 2 Map 124 Figure 53. Proportion of intersections in dimension 3 127 Figure 54. Dimension 3 Map 128 Figure 55. Districts Summary – Dimension 3 Map 129 Figure 56. The control loop 131 Figure 57. Clustering GET’s Map 132 LIST OF ANNEXES Annex C1. Administrative Division Map Annex C2. BTS and 4G Coverage Map Annex C3. Modernization and Maintenance Tender Map Annex C4. Road Hierarchy Map Annex C5. Management Areas Map Annex C6a. Traffic Light System Map Annex C6b. Traffic Light System Map – Pedestrian Annex C6c. Traffic Light System Map – Cyclist Annex C7. Controllers Map Annex C8. Controller Supplier Map Annex C9. Subway and Train Network Map Annex C10a. City Bus Network Map – Bus Corridors and Bus Exclusive Lanes Annex C10b. City Bus Network Map – Bus Lines Annex C11. Duct Network Map Annex C12. Cycling Network Map Annex C13. Dimension 01 Annex C14. Dimension 02 Annex C15. Dimension 03 Annex C16 - Districts Summary - Dimension 01 Map Annex C17 - Districts Summary - Dimension 02 Map Annex C18 - Districts Summary - Dimension 03 Map Annex C19 - Clustering GET's Map Annex C20.1 - Summary Clustering GET- CN Map Annex C20.2 - Summary Clustering GET- NO Map Annex C20.3 - Summary Clustering GET- SE Map Annex C20.4 - Summary Clustering GET- SU Map 8 of 134 C - Diagnosis Annex C20.5 - Summary Clustering GET- SO Map Annex C20.6 - Summary Clustering GET- MB Map Annex C20.7 - Summary Clustering GET- LE Map Annex C20.8 - Summary Clustering GET- OE Map LIST OF ACRONYMS Anatel – National Telecommunications Agency (Agência Nacional de Telecomunicações) AVL – Automatic Vehicle Location BTS – Base Transceiver Stations CADE – Administrative Council for Economic Defense (Conselho Administrativo De Defesa Econômica) CCO – Operational Control Centrals (Central de Controle Operacional) CEAGESP – São Paulo Warehouse and General Storage Company (Companhia de Entrepostos e Armazéns Gerais de São Paulo) CET – Traffic Engineering Company (Companhia de Engenharia de Tráfego) CETESB – Environmental Company of São Paulo State (Companhia Ambiental do Estado de São Paulo) CMCT – Municipal Company of Collective Transports (Companhia Municipal de Transportes Coletivos) COP – Operations Central (Centro de Operações da SPTrans) CPTM – São Paulo Metropolitan Train Company (Companhia Paulista de Trens Metropolitanos) CTA – Traffic Centrals by Area (Centrais de Tráfego em Área) CTB – Brazilian Traffic Code (Código de Trânsito Brasileiro) DCS – Traffic Light Control Department (Departamento de Controle Semafórico) DENATRAN – National Traffic Department (Departamento Nacional de Trânsito) DET – Traffic Engineering Department (Departamento de Engenharia de Tráfego) DSV – Road System Operation Department (Departamento de Operação do Sistema Viário) EMPLASA – São Paulo Company for Metropolitan Planning (Empresa Paulista de Planejamento Metropolitano) FTTH – Fiber to the Home FUMID – Municipal Fund for Digital Inclusion (Fundo Municipal de Inclusão Digital) GDP – Gross Domestic Product GET – Traffic Engineering Management (Gerência de Engenharia de Tráfego) GHG – Greenhouse gases GPON – Gigabit Passive Optical Network GPRS – General Packet Radio Services IBGE – Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística) ICT – Information and Communication Technology IML – Medico-Legal Institute (Instituto Médico Legal) ITS – Intelligent Transportation Systems LAN – Local Area Network 9 of 134 C - Diagnosis Metrô – São Paulo Subway Company (Companhia do Metropolitano de São Paulo) MNOs – Mobile Network Operators NTCIP – National Transportation Communications for ITS Protocol OD – Origin-Destination (Origem-Destino) PDUI – Integrated Urban Development Plan (Plano de Desenvolvimento Urbano Integrado) PEC – Emergency Plan for Sidewalks (Plano Emergencial de Calçadas) PL – Bill (Projeto de Lei) PMI – Procedure for Expression of Interest (Procedimento de Manifestação de Interesse) PMSP – São Paulo City Council (Prefeitura Municipal de São Paulo) PPI – Investment Partnerships Program (Programa de Parceria de Investimento) PPMI – Preliminary Procedure for Expression of Interest (Procedimento Preliminar de Manifestação de Interesse) RAIS – Annual Social Information Registration (Relação Anual de Informações Sociais) RMSP – Metropolitan Region of São Paulo (Região Metropolitana de São Paulo) SAT – Traffic Accident System (Sistema de Acidentes de Trânsito) SCOOT – Split Cycle and Offset Optimisation Technique Seade – State Data Analysis System Foundation (Fundação Sistema Estadual de Análise de Dados) SGM – Secretary of Municipal Government (Secretaria de Governo Municipal) SIM – Integrated Monitoring System (Sistema Integrado de Monitoramento) SLP – Private Limited Service (Serviço Limitado Privado) SLT – State Secretary of Logistics and Transport (Secretaria de Logística e Transportes) SMDP – Executive Secretary of Privatization and Partnerships (Secretaria Executiva de Desestatização e Parcerias) SMIT – Municipal Secretary of Innovation and Technology (Secretaria Municipal de Inovação e Tecnologia) SMS – Secretary of Municipal Health (Secretaria Municipal de Saúde) SMT – Municipal Secretary of Mobility and Traffic (Secretaria Municipal de Mobilidade e Trânsito) SMT* – State Secretary of Metropolitan Transports (Secretaria de Estado dos Transportes Metropolitanos) SPTrans – São Paulo Transport (São Paulo Transporte) SVMA – Municipal Secretary of Green and Environment (Secretaria Municipal do Verde e Meio Ambiente) TCU – Federal Court fo Accounts (Tribunal de Contas da União) TETRA – Terrestrial Trunked Radio UN – United Nations UPS – Uninterruptible Power Supply UTMC – Urban Traffic Management Control VLAN – Virtual Local Area Network VTR – Rapid Transit Road (Vias de Trânsito Rápido) WHO – World Health Organization 10 of 134 C - Diagnosis GLOSSARY 5G/NR The fifth generation technology standard for broadband cellular networks (5G) was presented for the first time as New Radio (NR) by the international norms and standardization organization 3GPP (3rd Generation Partnership Project) with the 3GPP Release-15. 5G is based on the existing Long Term Evolution (LTE) standard. The term 5G New Radio, abbreviated to 5G NR, stands for the air interface (radio interface) in fifth generation cellular networks (5G). 5G New Radio should enable high data rates in the double-digit gigabit range and ensure optimal spectral efficiency of the radio frequencies used. Further requirements for the technology are the support of large connection densities of up to one million mobile devices per square kilometer and short latency times of just one millisecond and less. In contrast to the third (3G; UMTS) or fourth generation (4G; LTE) mobile radio standards, 5G NR supports a much larger range of radio frequencies. Active mobility Active mobility represents non-motorized means of transportation that are based on human physical activity, such as walking, cycling and so on. Central Adaptive or Control mode in which, due to traffic conditions variation within a corridor network optimization or network, is necessary to do real-time adjustments. These through control decisions taken at a central level. These decisions are executed in the controller mostly by sending variables for green time, cycle time lengths, and offsets adjustments. There are also solutions that can include a dynamic signal program selection within an available predefined library of scenarios that are thought to attend expected demand variation. Control methods Urban control systems employ various traffic response methods that allow the basic control elements, cycle time, phase splits, offset, gaps measurement, stages on demand, to vary according to prevailing traffic conditions. The methods can be local or in the network. At the local level, algorithms are established and executed in the controller itself, such as Dual Ring & Barrier, VSPlus and Individual logic (i.e. LISA, Trelan-Trends, PDM). Regarding network control, the most common options are handled at the central level such as SCOOT, SCATS, InSync, INES, among others. Crashes Crashes are events that can be considered in terms of injuries, fatalities, or property damage [1]. For this study, ‘crashes’ will be a synonym for the Portuguese term ‘sinistros de trânsito’ that represents every event that results in damage to the vehicle or its cargo and/or injuries to people and/or animals, and that may cause material damage or harm to the traffic, road or to the environment, when at least one party is moving on roads or areas open to the public [2]. Fixed Times Control mode based on a scheduled signal plans programmed in the controller and designed off-line. 11 of 134 C - Diagnosis ITS Advanced applications aimed at offering innovative services related to different modes of transport and traffic management, enabling users to be better informed and make safer, more coordinated, and "smarter" use of mobility infrastructure and transport networks. MaaS Mobility as a Service. Type of service that through a joint digital channel enables users to plan, book, and pay for multiple types of mobility services. [3] MNO A mobile network operator (MNO) is a company that operates a public mobile network and offers services for private and business customers. For this purpose, the MNO applies for or acquires (e.g. by means of an auction) a broadcasting license from government agencies. The licensee then builds an infrastructure (today in the vast majority of cases a GSM, UMTS or LTE network) that covers a specific geographical area corresponding to the license. Mobile network operators must be differentiated from those who specialize in reselling mobile services without their own network ((Mobile Virtual Network Enablers and Mobile Virtual Network Operators). Neutral Host Network A neutral host network describes a special network architecture in which an independent, neutral third-party provider (e.g., a PPP, a management company, a private company, etc.) operates an expanding mobile communications infrastructure and makes it available to third parties as a business model. Open standard Are the kind of protocols that can be included in a wide range of device communication protocols types from any equipment vendor. This means that manufacturers that choose to adopt an open protocol want to achieve protocol interoperability when they design their equipment's functionality and capabilities. [4] Protocols Network protocols are a set of rules, conventions, and data structures that dictate how devices exchange data across networks. In other words, network protocols can be equated to languages that two devices must understand for seamless communication of information, regardless of their infrastructure and design disparities. [5] Smart Cities The concept of Smart Cities is defined by the use of technology to improve urban infrastructure and make urban centers more dynamic and efficient, which means also quality of life enhancement for its community. The interconnection of data facilitates the cross-referencing and analysis of information and provides a solid basis for the for the implementation of more effective public policies. When we talk about smart cities, we think of innovation, technology, sustainability and urban planning. [6] Traffic-Actuated Control mode based on a variable programming by means of a local algorithm uploaded in the traffic controller. Allows optimizing the 12 of 134 C - Diagnosis operation of the intersection according to the reading of the existing detectors connected to the intersection, in the frame of which is established in the control algorithm. It is also known as micro-regulation. 13 of 134 C - Diagnosis 1 INTRODUCTION Embedded in a number of other important ‘Smart City’-projects supported by the World Bank to improve the infrastructure in the city of São Paulo with the intention, not only to ease the daily life and social interaction of its inhabitants, but also to further advance economic and ecological development in a sustainable manner, this “Study on benchmarking and analysis of future technologies fo r the modernization, expansion and adaptation of the traffic lights systems in relation to the advent of 5G technologies.” bridges the gap between intelligent traffic management and new, pioneering telecommunications technology, which with 5G now holds out the prospect of completely new fields of application and solutions to long-standing problems. Our consortium ‘Future Mobility for São Paulo’, consisting of an interdisciplinary team of transport, telecommunications and financing experts, aims to collect, analyze and evaluate knowledge of current and coming technological alternatives and aims to discuss to what amount and in which circumstances 5G could be an enabler for new solutions and business models. First, we let our gaze wander widely and create a general overview. Developing criteria to identify the most promising developments and trends then allows us to narrow down the multitude of possible solutions to São Paulo's requirements and initial situation, and to recommend feasible implementation steps. We then broaden the perspective again and describe an ideal approach to individually address the different traffic management challenges in urban areas. Therefore, we present our results in various reports that have a common basis and build on each other: • We first talked to the different stakeholders of the project and captured their interests and goals and adjusted our Work Plan (A) accordingly. • A systematic Benchmark (B), where we directly interviewed different cities, technology manufacturers and also research institutes and universities worldwide, in addition to a metadata analysis of freely available information, e.g. websites and press releases, forms our knowledge base. Here, we first list trends & developments, products and services as well as different prerequisites, framework conditions and special challenges. We present the various solution approaches and visions neutrally next to each other. • In parallel, we survey, describe and study the situation in and the specific requirements of Sao Paulo in order to be able to provide a detailed Diagnosis (C). • In deliverable (D)-Recommended Technological Options, we establish criteria and KPI to evaluate the possible technological options found in (B). Listing and explaining arguments for or against a solution in defined scenarios leads us to general recommendations. • Adapting these to the specifics of São Paulo, we will be able to describe then in (E) the Options and Implications for the city in detail and step by step towards different levels. • Back to a more general view, we will then summarize all our findings and lessons learned from São Paulo in a transfer report to give (F) Final Recommendations to the World Bank. • The Dissemination (G) will consist of an Executive Summary and Workshops to present the results to the WorldBank. In addition to the requirements, we plan to edit a catalogue of summarized information in the form of “One Pagers” which will be distributed to all stakeholders and participants of our interviews with the goal to enable networking in between them. 14 of 134 C - Diagnosis F IGURE 1. S CHEMATIC CHRONOGRAM AND OVERVIEW OF REPORTS Benchmark (B) Stakeholder Recommended Options and Final Analysis & Dissemination Technological implications Recommendations adapted (G) Options (D) for SP (E) Worldwide (F) Workplan (A) Diagnosis/ Status quo São Paulo (C) Source: Future Mobility for São Paulo. Since the common thread running through all of our research must always be feasibility and cost- effectiveness that meets (local) regulatory requirements, we will also describe possible business models and ideas for how to finance the implementation and operation of each step in all reports. Due to the numerous options, we have chosen the strategy of defining a set of scenarios. Only by making assumptions and defining a context is it possible to effectively compare the costs and effort, risks and chances of a solution and to finally make a recommendation. Our scenarios are described in horizontal levels and vertical steps: The horizontal levels representing different views of/aspects of a scenario, e.g., equipment needed, means of data transfer, forms of management, etc. The vertical steps, on the other hand, describe the evolution from the status quo, starting up towards current state-of-the-art technology on to a future possibly wireless vision for intelligent adaptive traffic management. We also learned that some terms are used differently in different circumstances and countries, may have different meanings and scopes. Therefore, we decided to define them for this study in order to have a common starting point and basis for discussion. You can find all the definitions of terms and abbreviations used in the Acronyms & Glossary document. Now, the purpose of this document (C)-Diagnosis is to provide a comprehensive diagnosis of São Paulo’s traffic light system and telecommunication network regarding infrastructure, regulation, operation and future prospects. Through data collection and the construction of a traffic light geodatabase, it aims at the consolidation of a basic scenario that will be the starting point for future technological recommendations. Below, there is a flowchart (Figure 2) summarizing all the steps and topics of discussion that integrate this document. For C-Diagnosis, the first step was to stablish a general context for the city of São Paulo regarding telecommunication and trends in mobility. Then, several reports and official documents were analyzed to gather relevant information and to build a database for São Paulo’s traffic light system. Finally, the traffic light system was characterized according to several indicators, resulting in dimensions that will serve as a basis for future recommendations. 15 of 134 C - Diagnosis F IGURE 2. D IAGNOSIS F LOWCHART Source: Future Mobility for São Paulo. 16 of 134 C - Diagnosis 2 GENERAL DATA 2.1 CITY OF SÃO PAULO São Paulo is the capital of the homonymous state, located in Brazil’s Southeast region. With a population of approximately 12.3 million1 people, São Paulo is the most populous city in the country and one of the few megacities2 in the world. According to IBGE (Brazilian Institute of Geography and Statistics), in 2018 São Paulo had the highest national GDP per capita and the highest national average wage of formal workers. About 90% of all generated product was by the tertiary sector (commerce and services). Here follows some relevant data about the city: ▪ Estimated population (IBGE): 12,325,232 inhabitants (2020) ▪ Area of territorial unit (IBGE): 1,521.11 km² (2020) ▪ Demographic density: 8,102.79 inhab/km² (2020) ▪ Human Development Index (IBGE): 0.805 (2010) ▪ GDP per capita (IBGE): BRL 58,691.90 (2018) ▪ City of São Paulo GDP (Seade): BRL 714,683,362,463.00 (2018) ▪ Average wage of formal workers (IBGE): 4.3 minimum wages3 (2018) ▪ Geographical coordinates (WGS-84): latitude 23º 29’ 27.92650" S and longitude 46º 36’ 54.49652" W ▪ Altitude (WGS-84): 762.406 meters Regarding its administrative division, the city of São Paulo has 96 districts distributed in 32 sub-prefectures (Annex C1. Administrative Division Map and Figure 3). The sub-prefectures were established in order to locally manage municipal affairs, carrying out urban maintenance actions and establishing a direct communication channel between government and population. 2.2 INSERTION IN THE METROPOLITAN REGION According to UN “World Urbanization Prospects 2018” report [7], the Metropolitan Region of São Paulo (SPMR) is the world’s fourth largest urban agglomeration, only behind Tokyo, Delhi, and Shanghai. São Paulo is considered the center of this metropolitan complex and corresponds to 58% of the SPMR’s population and 60% of the total GDP. Here follows some information about the Metropolitan Region of São Paulo: ▪ Estimated population (IBGE): 21,120,832 inhabitants (2020) ▪ Area of territorial unit (PDUI): 7,946.96 km² ▪ Demographic density: 2,657.72 inhab/km² (2020) ▪ Metropolitan Region GDP (Seade): BRL 1,181,500,892,256.00 (2018) 1 The estimated population of São Paulo for 2020 is 12,325,232 inhabitants, a number calculated by IBGE (Brazilian Institute of Geography and Statistics) based on the AiBi mathematical method. 2 According to “World Urbanization Prospects 2018” [7] report by the United Nations Department of Economic and Social Affairs, megacities are urban agglomerations with 10 million inhabitants or more and have a high concentration of economic activities. 3 In 2018, the minimum wage in Brazil was BRL 954.00. Thus, the average wage in São Paulo in this period corresponds to BRL 4,102.20, equivalent to 4.3 minimum wages. 17 of 134 C - Diagnosis F IGURE 3. A DMINISTRATIVE DIVISION MAP Source: Future Mobility for São Paulo. 18 of 134 C - Diagnosis Composed of a continuous urban area of 39 municipalities, the SPMR has great political and economic influence in the State of São Paulo and in Brazil as a whole, concentrating 46% of the State’s population and 17% of Brazil’s GDP. Its economic profile is quite diversified, with an emphasis on the service sector in areas such as telecommunications, culture, education, health, transport, and gastronomy. In addition, SPMR can be considered the greatest center of national wealth as it contains a few of the most important industrial parks in Brazil – São Paulo, ABC Region, Guarulhos, and Osasco – and the most important financial center in Latin America – the B3 Stock Exchange [8]. The SPMR’s municipalities are divided into 5 sub-regions (except São Paulo, which is included in all sub- regions): ▪ North: Caieras, Cajamar, Francisco Morato, Franco da Rocha e Mairiporã ▪ East: Arujá, Biritiba-Mirim, Ferraz de Vasconcelos, Guararema, Guarulhos, Itaquaquecetuba, Mogi das Cruzes, Poá, Salesópolis, Santa Isabel e Suzano ▪ Southeast: Diadema, Mauá, Ribeirão Pires, Rio Grande da Serra, Santo André, São Bernardo do Campo e São Caetano do Sul ▪ Southwest: Cotia, Embu, Embu-Guaçu, Itapecerica da Serra, Juquitiba, São Lourenço da Serra, Taboão da Serra e Vargem Grande Paulista ▪ West: Barueri, Carapicuíba, Itapevi, Jandira, Osasco, Pirapora do Bom Jesus e Santana de Parnaíba Each municipality has an independent administration, with the PDUI (Integrated Urban Development Plan) as a planning and integration tool for urban public policies in the metropolitan region. The SPMR’s PDUI was approved in April 2019, contemplating four functional development axes that correspond to the regional structural problems: ▪ Economic, Social and Territorial Development ▪ Housing and Social Vulnerability ▪ Environment, Sanitation and Water Resources ▪ Mobility, Transport and Logistics 2.3 ORGANIZATION OF PUBLIC AND PRIVATE SECTOR 2.3.1 São Paulo State Government São Paulo State is one of the 27 federal units in Brazil – composed of 26 states and 1 federal district. Each state is divided into municipalities, São Paulo has a total of 645 municipalities including the city of São Paulo which is its capital. State and municipal governments have different responsibilities and funding sources. For example, the state of São Paulo manages metropolitan train and intercity bus lines, while the city of São Paulo manages city bus lines. To organize public initiatives from multiple sectors, there are 27 secretaries that integrate São Paulo State Government. From these, two are related to transports: the State Secretary of Logistics and Transport (SLT) and the State Secretary of Metropolitan Transports (SMT*). 2.3.2 Metrô – São Paulo Subway Company The São Paulo Subway Company – also known as Metrô – was created in 1968 and is responsible for the operation, expansion and planning of SPMR’s subway network. Currently, it is controlled by São Paulo State Government under the State Secretary of Metropolitan Transports. For more details about Metrô’s infrastructure and operation, see item 7.1.2.1 of this report. 2.3.3 CPTM – São Paulo Metropolitan Train Company The São Paulo Metropolitan Train Company (CPTM) was created in 1992 to manage the existing train system in São Paulo Metropolitan Region and optimize its operation alongside Metrô. Currently, it is 19 of 134 C - Diagnosis controlled by São Paulo State Government under the State Secretary of Metropolitan Transports. For more details about CPTM’s infrastructure and operation, see item 7.1.2.2 of this report. 2.3.4 PMSP – São Paulo City Hall São Paulo Municipal Government is composed of 28 secretaries, each secretary responsible for a determined area of public administration, such as: health, culture, education, transports, international relations and so on. In addition to secretaries, which work directly with the Mayor, São Paulo City Hall has many indirect administration entities and public companies that manage different public services. 2.3.5 SMT – Municipal Secretary of Mobility and Traffic SMT is the municipal secretary responsible for managing municipal services related to urban mobility and traffic. It was created in 1967 as Municipal Secretary of Transports by Law nº 7.065/67, and now is organized by Decree nº 60.448/21. This last decree altered the secretary name to ‘Mobility and Traffic’ and assigned the following duties: ▪ Formulate, propose, manage, and evaluate public policies for the development of sustainable, integrated, and efficient urban mobility, while prioritizing the preservation of life, health, and the environment. ▪ Regulate and inspect the use of municipal network of roads and bike lanes. ▪ Regulate, manage, integrate, and supervise collective and individual means of transport of people and cargo, motorized and active, including school transport, within its competence. ▪ Encourage active transportation and micro mobility associated with low environmental impact propulsion integrated to the road network. ▪ Plan and execute traffic services and traffic control within its competence, as well as promote traffic education and safety. ▪ Execute activities that are compatible and related to its area of expertise. SMT has two associated indirect administration entities: CET and SPTrans. The secretary is responsible for managing these two contracts and supervising their performance, as well as integrating studies related to their competence. 2.3.6 CET – Traffic Engineering Company CET is a company created in 1976 – in which São Paulo City Hall is a majority shareholder – responsible for the city traffic management. It is an indirect administration entity hired by SMT to carry out the following activities: ▪ Traffic operation: inspection by managers and technicians; operational central for monitoring occurrences; violation inspection by traffic agents; implementation of radars and cameras; traffic light control centrals; management of “Zona Azul”4; and transportation of large and dangerous products in the city. ▪ Planning studies and projects: expansion and improvement of road network; signaling project manuals; active mobility and public transport studies; project of complex intersections; traffic detour for civil works and so on. 4 “Zona Azul” are delimited parking spaces in the city in which the user needs to pay a fee (per hour). Currently, the operation of this service is carried out by Estapar, a private company that recently won a public bidding for 15 years of concession. 20 of 134 C - Diagnosis ▪ Road safety: accident data survey; studies or analysis for places where accidents occur; monitoring and dissemination of results. ▪ Implementation and maintenance of traffic signaling road signs; road marking; traffic lights; and traffic central for monitoring. ▪ Traffic education and training: lectures and campaigns for schoolchildren, college students, teachers, cyclists, bikers, taxi drivers and car drivers. ▪ Strategic studies: legislation studies; studies on car use discouragement; elaboration of action policies; participation in congresses and so on. ▪ Development of new technologies: development of new road signs; radars; and traffic control centrals. For more information about CET’s traffic operation in the city of São Paulo, especially regarding traffic light infrastructure, see item 7.1 of this report. 2.3.7 SPTrans – São Paulo Transport SPTrans is an indirect administration entity responsible for managing the operation of city buses in São Paulo, for city bus details see item 7.1.2.3 of this report. It was named São Paulo Transport SA in 1995, substituting CMCT – Municipal Company of Collective Transports (Companhia Municipal de Transportes Coletivos) – created in 1946. According to SPTrans bylaws, for the operation of collective public transportation in São Paulo it may take on the following activities: ▪ Open biddings, sign contracts, and grant permissions regarding service provision of public transport. ▪ Elaborate technical, economic, and financial studies to plan and improve the system. ▪ Manage and supervise the service provision of public transport implementation, improvement, administration, and expansion. For this, the use of financial and budgetary resources must be in accordance with PMSP and SMT’s guidelines. ▪ Manage resources from fare collection and fining of hired companies. ▪ Manage and supervise the service provision of hired companies related to public transport. ▪ Apply penalties for infractions related to service provision of public transport. 2.3.8 SMIT – Municipal Secretary of Innovation and Technology SMIT is the municipal secretary responsible for promoting innovation and technology in the services provided by public administration. It was created in 2018 by Law nº 16.974/18 and was recently reorganized by Decree nº 59.336/2020, which assigned the following duties: ▪ Promote improvement and innovation in the organization and services provided by Municipal Public Administration, in order to increase service quality and promote citizen participation in the development of a smart and human city. ▪ Promote digital inclusion, access to information and ICT (Information and Communication Technology), in order to expand digital citizenship. ▪ Disseminate the use of technologies that contribute to urban economic development, especially in areas of greater social vulnerability. ▪ Encourage the development of ICT management within public administration, in order to establish conditions for the construction of a Digital Government. ▪ Manage the Municipal Fund for Digital Inclusion – FUMID (Fundo Municipal de Inclusão Digital). Among its services, it is possible to highlight the WiFi Livre SP initiative, which provides public internet access though hotspots located in public facilities. Currently, there are 149 access points located in 21 of 134 C - Diagnosis cultural centers, libraries, parks, and schools. Also, SMIT hosts an Innovation Hub responsible for connecting different stakeholders (private companies, universities, research institutions, and so on) who wish to collaborate with São Paulo City Hall, as well as promoting technological projects in SMIT and other municipal secretaries5. 2.3.9 Anatel – National Telecommunications Agency Anatel is a federal regulatory agency created in 1997 by Law nº 9.472/97, also known as the General Law of Telecommunications. It is an indirect administration entity of the Federal Government responsible for adopting necessary measures to ensure that public interests are met, as well as the development of Brazilian telecommunication sector. Here follows some of Anatel’s duties: ▪ Implement a national policy for telecommunication. ▪ Represent Brazil at international telecommunication organizations. ▪ Manage radio frequency spectrum and orbit use, issuing the respective regulation. ▪ Issue or recognize product certification, observing established regulation and standards. ▪ Manage conflicts of interest between telecommunication service providers. ▪ Restrain violation of users’ rights. ▪ Exercise, regarding telecommunications, legal power in matters of control, prevention and repression of economic infractions, except those belonging to Administrative Council for Economic Defense (CADE). 2.3.10 MNOs – Mobile Network Operators São Paulo has many MNOs operating in the city to provide mobile phone services. To analyze the relevance of each company, market share (Figure 4) and number of accesses (Figure 5) were taken into consideration. F IGURE 4. MARKET SHARE OF MOBILE PHONE SERVICES IN THE CITY OF SÃO P AULO Datora Outros 1.70% 1.20% Oi 10.80% Vivo 35.60% Tim 18.90% Claro 31.80% Source: Anatel (June 2021), https://informacoes.anatel.gov.br/paineis/acessos. 5 Information provided during an internal meeting with SMIT in August 2021. 22 of 134 C - Diagnosis F IGURE 5. NUMBER OF MOBILE PHONE ACCESSES IN THE CITY OF SÃO PAULO Vivo 8,653,397 Claro 7,734,666 Tim 4,606,241 Oi 2,615,316 Datora 402,837 J. Safra Telecomunicações 229,831 Surf Telecom S.A. 58,498 America Net S.A. 12,311 0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 Source: Anatel (June 2021), https://informacoes.anatel.gov.br/paineis/acessos. Vivo, Claro, Tim and Oi are considered ‘big companies’, and together they sum 97.10% of mobile phone market share. In 2016, Oi filed for judicial recovery due to an accumulation of debts and has since been selling its assets. The remaining ‘big companies’ created a consortium and, in December 2020, bought Oi’s mobile network for BRL 16.5 billion – this operation is still waiting for the approval of CADE. On the other hand, companies that own less than 5% of market share are considered ‘small-scale companies’ by Anatel. Those are: Datora, J. Safra Telecomunicações, Surf Telecom S.A. and America Net S.A. 23 of 134 C - Diagnosis 3 TELECOMMUNICATION IN BRAZIL During a long time, Telecommunications in Brazil was operated by Union controlled companies, one per state, and one responsible for national and international traffic. On July 16th 1997, the General Law of telecommunications was issued in Brazil. This Law means to regulate how telecommunications must be in Brazil. Art. 1 It is incumbent upon the Union, through the regulatory body and under the terms of the policies established by the Executive and Legislative Powers, to organize the exploitation of telecommunications services. Single paragraph. The organization includes, among other aspects, the discipline and inspection of the execution, commercialization and use of services and the implementation and operation of telecommunications networks, as well as the use of orbit and radio frequency spectrum resources. The article 6, below, says that all telecommunications will be offered by Providers observing the free, broad, and fair competitions: Art. 6 The telecommunications services will be organized based on the principle of free, broad, and fair competition between all providers, and the Government shall act to provide it, as well as to correct the effects of imperfect competition and repress infringements of the economic order. Telecommunications in Brazil is only via Providers: Article 3 – Users have following rights: 1. Can choose the provider as his own will 2. Cannot be discriminated on access conditions and services Article 8 – Creates ANATEL – National Telecommunications Agency Responsible to regulate telecommunications activities in Brazil. ANATEL is also responsible for the administration of radio frequencies spectrum and satellite orbits. All infrastructure (ducts, cables, towers) is made and maintained by the Provider. All services are offered by the Providers. Claro, Vivo and OI are the major internet wide band Providers in Brazil, together they own 57.40% of the market share as it is shown in Figure 6. All other Providers have less than 2% of market share each. 24 of 134 C - Diagnosis F IGURE 6. Q UANTITY OF INTERNET WIDE BAND ACCESSES BY PROVIDER IN BRAZIL CLARO 9,774,338 VIVO 6,337,385 OI 5,205,908 Wide Band Providers Brisanet 736,580 ALGAR 726,985 TIM 672,980 UNIFIQUE 370,785 DESKTOP-SIGMANET 333,158 SUMICITY 324,055 AMERICA NET 315,037 Others 14,316,352 0 4,000,000 8,000,000 12,000,000 16,000,000 Quantity of Accesses Source: Anatel (June 2021), https://informacoes.anatel.gov.br/paineis/acessos/ranking. For Mobile Phone services, Claro, Vivo, TIM and OI are the major Providers in Brazil. Together they own 97.96% of the market share as it is shown in Figure 7. F IGURE 7. Q UANTITY OF MOBILE PHONE ACCESSES BY PROVIDER IN B RAZIL VIVO 80,965,123 CLARO 67,774,992 TIM 51,340,682 Mobile Phone Providers OI 40,333,376 ALGAR 2,962,746 DATORA 880,976 SURF TELECOM 684,053 J. SAFRA 356,439 AMERICA NET 86,850 COPEL 46,807 LIGUE MOVEL 79 0 20,000,000 40,000,000 60,000,000 80,000,000 100,000,000 Quantity of Accesses Source: Anatel (June 2021), https://informacoes.anatel.gov.br/paineis/acessos/ranking. On the other hand – for cable TV –, Claro, Sky/AT&T, OI, and Vivo are the major Providers in Brazil and together they own 97.45% of the market share (Figure 8). Claro alone represents 47% of the total number of accesses. 25 of 134 C - Diagnosis F IGURE 8. Q UANTITY OF CABLE TV ACCESSES BY PROVIDER IN BRAZIL CLARO 6,531,135 SKY/AT&T 4,080,603 OI 1,757,172 VIVO 1,186,241 Cable TV Providers Brisanet 39,544 NOSSATV 34,000 CABO 28,297 OPCAONET 19,254 VIDEOMAR 18,160 RBC 18,144 Others 197,534 0 1,000,000 2,000,000 3,000,000 4,000,000 5,000,000 6,000,000 7,000,000 Quantity of Accesses Source: Anatel (June 2021), https://informacoes.anatel.gov.br/paineis/acessos/ranking. At last, OI, Claro, and Vivo are the major fixed telephony Providers in Brazil and together they own 87.51% of the market share as it is shown in Figure 9. F IGURE 9. Q UANTITY OF FIXED TELEPHONY ACCESSES BY PROVIDER IN BRAZIL OI 9,121,646 CLARO 8,870,134 VIVO 8,301,240 Fixed Telephony Providers ALGAR 1,289,517 TIM 836,647 AMERICA NET 280,862 UNIVERSO 271,754 SUMICITY 256,888 COPEL 208,669 DESKTOP 94,080 Others 514685 0 2000000 4000000 6000000 8000000 10000000 Quantity of Accesses Source: Anatel (June 2021), https://informacoes.anatel.gov.br/paineis/acessos/ranking. 3.1 5G AUCTION IN BRAZIL In November 2019, the Investment Partnerships Program (PPI) Council – initiative created to facilitate interactions between private companies and Public Power for contracts and other privatization measures 26 of 134 C - Diagnosis – qualified the 5G Auction in Brazil through Resolution nº 88/2019. After many delays, the 5G Auction Tender was approved by Anatel’s board of directors in February 2021, and later validated by the Federal Court of Accounts (TCU) in August 2021. Expected to be the biggest spectrum offer in Anatel’s history, the frequency bidding includes 700MHz, 2.3 GHz, 3.5 GHz, and 26GHz bands with a commercial exploitation period of 20 years. On September 24th, 2021, Anatel approved the 5G Auction Tender final version and defined that interested companies can start sending their documentation on October 27 th. Price proposals will be analyzed in November 4th, 2021 [9]. Discussions about a new mobile communication bidding for the 700 MHz frequency band started in early 2018, which evolved to include the 2.3 GHz, 3.5 GHz, and 26 GHz bands in 2019. These spectrums were consolidated by Ordinance nº 1.924/SEI-MCOM, document published by the Federal Ministry of Communications in January 2021 to provide guidelines for future biddings of the forementioned frequency bands, such as: ▪ Encourage sharing of active and passive infrastructure among providers, including poles, towers, and ducts. ▪ Establish coverage commitments, including: provide mobile broadband access (4G or higher) to cities, villages, isolated urban or rural areas that have over 600 habitants; provide mobile broadband access in Federal highways; and implement optical fiber in cities with no backhaul infrastructure. ▪ Enforce the creation of a private network for Brazil Federal Government, including: mobile network in frequency band 703MHz to 708 MHz and 758MHz and 763 MHz for urgent Federal matters and emergencies; and a fixed network for Federal public institutions. ▪ Establish measures to solve interference issues in the 3.5 GHz frequency band caused by satellite television services – even if the solution is to digitalize and migrate this service to another frequency band. ▪ Encourage the use of open access network in order to promote interoperability between equipment of multiple suppliers. Regarding the last topic, Anatel6 stated there is an internal study group led by the Superintendence of Authorization and Spectrum Management that is researching interoperability issues to facilitate interconnection between networks and improve quality of service in a 5G environment. In this context, Anatel will regulate technical specifications that must be followed by equipment suppliers and manufacturers. However, it is not Anatel’s attribution to define or restrict specific companies that will provide hardware and software. For applications in 5G, Anatel’s understanding is that it could extrapolate general service provision – mobile phone networks, for example – and be destined to specific groups through a Private Limited Service (SLP) authorization. A SLP is a restricted telecommunication service regulated by Anatel’s Resolution nº 617/2013 that can operate in 5G frequencies not included in this first auction. The creation of a restricted network by this mechanism could be used for traffic management in urban areas, for instance. 6 Information provided during an internal meeting with Anatel in July 2021. 27 of 134 C - Diagnosis As for coverage, the 5G Auction will not specify priority areas or a minimum percentage of 5G coverage for cities, only the implementation of a certain number of stations. Therefore, if a company decides to install all 5G stations in a small portion of the city, technically, it is complying with the rules established by Anatel. It is expected that MNOs, to operate a comprehensive mobile communication service, will cover most regions in large cities. However, there is no obligation to do so, which could implicate in a market-oriented distribution of technology in detriment of an accessible and equitable network for collective use – such as traffic management. Finally, regarding implementation chronogram7, the 5G Auction Tender demands that each Brazilian State capital must have 5G in operation until July 2022. Municipalities with less than 30 thousand inhabitants must be covered until July 2030. The minimum standard for 5G operation must follow the general regulation for service quality established by Anatel (Resolution nº 717, from December 2019), as well as complementary norms that are still under development. In addition, strategies for 5G implementation in Brazil – e.g., non-stand-alone or Dynamic Spectrum Sharing (DSS) – will be defined by the MNOs that acquired frequency bands in the Auction. For instance, currently there are some areas in Brazil with access to 5G DSS, which uses part of the spectrum from other technologies (like 4G). More details about deadlines and requirements to be met by winning MNOs are shown in Table 1. T ABLE 1. 5G IMPLEMENTATION PLAN IN BRAZIL Deadline 5G implementation requirements State capitals and Federal District (1 station for each 100 July/2022 thousand inhabitants). State capitals and Federal District (1 station for each 50 July/2023 thousand inhabitants). State capitals and Federal District (1 station for each 30 July/2024 thousand inhabitants). State capitals and Federal District (1 station for each 15 thousand inhabitants). July/2025 Municipalities with more than 500 thousand inhabitants (1 station for each 15 thousand inhabitants). Municipalities with more than 200 thousand inhabitants July/2026 (1 station for each 15 thousand inhabitants). Municipalities with more than 100 thousand inhabitants July/2027 (1 station for each 15 thousand inhabitants). 50% of municipalities with population between 30 and July/2028 100 thousand inhabitants (1 station for each 15 thousand inhabitants). 7 Information provided by Anatel in November 2021. 28 of 134 C - Diagnosis Deadline 5G implementation requirements 100% of municipalities with population between 30 and July/2029 100 thousand inhabitants (1 station for each 15 thousand inhabitants). A fraction of municipalities with population under 30 July/2030 thousand inhabitants (quantity to be defined by the value of frequency band). Source: Anatel, 2021. Even with the deadlines for 5G implementation laid out, there is no forecast for shutting down existing mobile phones standards (like 3G). Anatel’s planning considers that Brazil’s mobile coverage will naturally evolve to 4G and 5G, according to established goals. In November 2021, Anatel analyzed the price proposals of 15 companies and/or consortiums for the 700MHz, 2.3GHz, 3.5 GHz and 26 GHz frequency bands being auctioned. From these, only 5 already operated telecommunications in Brazil: Vivo, Claro, TIM, Algar Telecom and Sercomtel. A total of 85% of the frequency bands offered were actioned off, raising an amount of BRL 47.2 billion (or 8.4 billion dollars) – of this value, BRL 39.8 billion will be reinvested in coverage expansion and telecommunication infrastructure [10]. The list of winners and permit values paid are available in Table 2. T ABLE 2. 5G AUCTION RESULTS AND PERMIT VALUES Frequency band Winning company Permit value 700MHz Winity II Telecom BRL 1.4 billion Claro BRL 338 million 3.5 GHz (National) Vivo BRL 420 million TIM BRL 351 million Sercomtel – North Region and São Paulo State Brisanet – Northeast and Midwest Regions 5G Sul Consortium – South Region 3.5 GHz (Regional) BRL 1.9 billion Cloud2U – Rio de Janeiro, Espírito Santo, and Minas Gerais States Algar Telecom – Some parts of Minas Gerais, Mato Grosso do Sul, Goiás and São Paulo States Claro – North, Midwest, and South Regions, plus São Paulo State 2.3 GHz (50MHz) Brisanet – Northeast Region BRL 1.5 billion Vivo – Rio de Janeiro, Espírito Santo, and Minas Gerais States Vivo – North and Midwest Regions, 2.3 GHz (40MHz) BRL 891 million plus São Paulo State 29 of 134 C - Diagnosis Frequency band Winning company Permit value TIM – South Region, Rio de Janeiro, Espírito Santo, and Minas Gerais States Algar Telecom – Some parts of Minas Gerais, Mato Grosso do Sul, Goiás and São Paulo States Claro (400 MHz – 20 years) 26 GHz (National) Vivo (600 MHz – 20 years) BRL 291 million TIM (200 MHz – 10 years) TIM – South and Southeast Regions Neko – São Paulo State Algar Telecom – Some cities in São 26 GHz (Regional) Paulo, Minas Gerais and Espírito BRL 61 million Santo States Fly Link – Some cities in São Paulo, Minas Gerais and Espírito Santo States Source: Anatel, 2021. As a result, Brazil now has new telecommunication service providers: Flylink, Neko, Brisanet, Cloud2U, 5G Sul Consortium, and Winity II Telecom. 3.2 NEUTRAL HOST A neutral host network describes a special network architecture in which an independent, neutral third- party provider (e.g., a PPP, a management company, a private company, etc.) operates an expanding mobile communications infrastructure and makes it available to third parties as a business model. 3.2.1 What is a Neutral Host Network? With increasing network densification, i.e., the transition from mobile networks based on macrocells with ranges of several kilometers to micro- or small-cell solutions with a few meters of network coverage, the number of hardware components in urban areas is also increasing. So far, each MNO has built its own network infrastructure in the urban area, which would lead to greater redundancy when small cells are deployed. Therefore, a shared infrastructure is beneficial to participate in lower installation heights from advantageous locations. These sites can be municipally or publicly owned, such as lampposts or traffic light poles, as well as privately owned, such as overhead poles or transit stops, and implemented as part of a neutral-host network. With the help of this open-access model, MNOs can buy capacity in the third-party provider's network under fair competitive conditions. In this case, it should be noted that the third-party provider must have appropriate frequency licenses in order to be able to offer its services in the licensed frequency bands. On the other hand, unlicensed frequency bands (e.g., WLAN frequencies) can also be used, although higher levels of interference from public and private WLAN networks can occur here and strong interference can therefore be expected. 30 of 134 C - Diagnosis Alternatively, framework agreements could be concluded with the public MNOs that allow services to be offered in the licensed frequency bands. Approaches for the implementation of neutral host networks already exist. These are planned in Germany, for example, for the provision of mobile communications services. MNO 1 uses the network architecture of MNO 2, but separate networks are provided instead of roaming or a national roaming. The implementation of such Neutral Host scenarios in traffic signal systems is described in Task B – International Benchmarking. The main advantages and disadvantages of a neutral host scenario are listed in Table 3: T ABLE 3. P ROS AND CONS OF A NEUTRAL H OST SCENARIO Pros Cons ▪ Fast rollout of new MNO. ▪ MNO circumvents obligation of ▪ Network densification of existing independent network expansion by networks (closing of unserved zones). cooperating with other MNOs. ▪ Cost-effective expansion of networks ▪ Failure of local infrastructure due to and improvement of coverage. malfunction, vandalism, etc. affects ▪ Saving of disruptive transmission towers multiple MNOs simultaneously. and infrastructure. ▪ Capacity limited due to use of parallel ▪ Flexible resource allocation according to MNOs, bottleneck in network bandwidth local demand for services from MNO 1 and performance. or MNO 2 etc. Source: Future Mobility for São Paulo Consortium. 3.2.2 Related Regulation in Brazil Regarding Neutral Host models in Brazil, shared infrastructure and network regulations were identified as a basis for future recommendations. In order to stimulate resource optimization and operational cost reduction, Anatel published Resolution nº 683/2017 to regulate shared infrastructure for telecommunication services. It establishes that infrastructure owners are obligated to share excess capacity for service providers when requested. Based on Anatel’s database for authorized BTSs (Base Transceiver Stations) in the city of São Paulo, there are over 900 sites with more than one station – which means these infrastructures are shared with more than one MNO. This model became more common as MNOs began to focus their capital on service provision rather than building and owning infrastructure. In Brazil, there are companies specialized in shared infrastructure – constructing and renting towers, for example –, such as American Tower do Brasil, Grupo TorreSur, SBA Communications, and QMC Telecom. These companies integrate ABRINTEL, an association formed by enterprises that own support infrastructure for indoor or outdoor BTSs, and whose services are directed to the expansion of telecommunications. Currently, companies associated with ABRINTEL represent 65% of the tower market (42,000 towers) in Brazil, having invested over BRL 15 billion in recent years [11]. 31 of 134 C - Diagnosis As for shared networks, Anatel’s Resolution nº 671/2016 regulates radiofrequency spectrum use – general parameters for administration, terms of use, authorization, and control of radiofrequencies in Brazil. It establishes that spectrum sharing – or “industrial exploration of radiofrequencies” – can be allowed as long as the positive effects outweigh negative ones. For this, at least one provider must be previously authorized to operate in the original frequency band. However, spectrum sharing still is a very bureaucratic process in Brazil, as it needs Anatel’s and CADE’s (Administrative Council for Economic Defense) approval. As an example, there are a few recent cases of RAN sharing contracts between MNOs operating in Brazil. In August 2021, Anatel approved Claro and Vivo 3G spectrum sharing in addition to 81 BTSs [12] – mostly located in small cities and highways. It configures a unilateral contract in which Claro will use Vivo’s network to improve telecom services in less populated regions. Another case is Vivo and TIM’s contract for 2G, 3G, and 4G network sharing, established in 2019. For 2G technology, this initiative will cover 2,700 cities and result in the deactivation of repeated sites for cost reduction and optimal spectrum use. Moreover, the contract aims to create a 3G and 4G single grid for small cities – under 30 thousand habitants [13]. Both of these operations are already in process of implementation [14] [15]. According to Anatel8, there are no obstacles in their current legislation to the existence of Neutral Host Networks for both mobile and fixed networks. 3.3 TELECOMMUNICATIONS IN SÃO PAULO According to Anatel’s dashboard for infrastructure data, São Paulo currently has 87.28% of its territory and 99.78% of its habitants covered by 4G technology (Annex C2. BTS and 4G Coverage Map). Considering all technologies – 2G, 3G, and 4G – these percentages rise to 89.27% and 99.82%, respectively. For mobile networks, Anatel defines that mobile communication services must cover at least 80% of a city’s urban area. Therefore, as it is shown in Table 4, São Paulo is considered covered. There are only 10 districts in São Paulo that contain areas with less than 95% of its territory covered by 4G, all located in the outskirts (extreme north, south and east zones): Anhanguera, Jaraguá, Brasilândia, Cachoeirinha, Mandaqui, Tremembé, Iguatemi, Parelheiros, Grajaú, and Marsilac districts. T ABLE 4. M OBILE NETWORK COVERAGE IN S ÃO PAULO Covered area Covered City MNO Technology (%) habitants (%) 4G 87.28% 99.78% 3G 84.55% 99.75% São Paulo All 2G 83.63% 99.68% All 89.27% 99.82% Source: Anatel (2021), https://informacoes.anatel.gov.br/paineis/infraestrutura. 8 Information provided by Anatel in November 2021. 32 of 134 C - Diagnosis Anatel also publishes a list of authorized BTSs (Base Transceiver Stations) in each city, with information about identification (ID number and name), location (address and geographic coordinates), and registration dates. Based on this list, São Paulo’s BTS netwo rk was georeferenced and accounted for (Annex C2. BTS and 4G Coverage Map and Figure 10). In August 2021, São Paulo had 6,110 BTSs in operation for four MNOs – Vivo, Claro, Oi, and TIM –, which represents an average of approximately 2,029 habitants per station. T ABLE 5. Q UANTITY OF BTSS (BASE TRANSCEIVER STATIONS ) PER DISTRICT – TOP 5 AND BOTTOM 5 Position District Quantity of BTSs 1º Itaim Bibi 485 2º Jardim Paulista 242 3º Santo Amaro 209 4º Pinheiros 201 5º Moema 196 92º São Miguel 18 93º José Bonifácio 15 94º Jaguará 14 95º Jardim Helena 14 96º Marsilac 4 Source: Anatel (August 2021). As for municipal regulation, São Paulo City Hall is responsible for the urban legislation and zoning, which directly affects BTSs’ location and approval. Ahead of 5G advent in Brazil, Anatel recently published guidelines [16] to encourage local governments to update their current legislation and reduce obstacles for the implementation of antennas. Along the same lines, in September 2020 a Presidential Decree (nº 10.480/2020) was published to stimulate the development of telecommunication infrastructure and regulate Law nº 13.116/2015, also known as the “General Law of Antennas”. According to Anatel, 5G technology will require a higher number of antennas to enable optimal coverage and transmission rates. Thus, cost reduction and simplification of administrative procedures – measures aligned with Law nº 13.116/2015 and Decree 10.480/2020 – are relevant initiatives to accelerate 5G implementation in Brazil. In this context, São Paulo City Hall is already mobilizing efforts to approve a new “Antenna’s Law” (PL 347/2021) to facilitate approval and installation of telecommunication infrastructure. In a letter attached to the bill, São Paulo’s Mayor Ricardo Nunes stresses that mobile coverage expansion is a matter of public interest by exemplifying existing socioeconomical inequalities and the lack of internet access in peripheral regions – situation aggravated by Covid-19 pandemic and the dissemination of solutions like digital classes or remote health services. São Paulo’s current “Antenna’s Law” (Law nº 13.756/2004) is considered very restrictive and was deemed unconstitutional by the Federal Supreme Court, as it invades the exclusive Federal Government competence to legislate telecommunications. As an example, Law nº 13.756/2004 establishes a minimum distance of 100 meters between towers, poles, and similar infrastructures – even when they are shared. Therefore, PL 347/2021 should be a key factor for 5G implementation in São Paulo since there are no 33 of 134 C - Diagnosis limitations for site location. In addition, the proposed law adopts a simplified process to implement BTSs in public equipment, such as: ▪ Road infrastructure (tunnels, bridges and so on). ▪ Street furniture. ▪ Street lighting poles. ▪ Traffic monitoring cameras. ▪ Surveillance cameras. ▪ Other equipment. 34 of 134 C - Diagnosis F IGURE 10. BTS AND 4G C OVERAGE MAP Source: Future Mobility for São Paulo. 35 of 134 C - Diagnosis Stations or mini stations installed in public equipment will pay a monthly fee for São Paulo City Hall, value and terms of use will be defined in a future regulation. Another relevant topic is the creation of priority areas to equalize telecommunication coverage in São Paulo. According to the proposed law, to build a new BTS in a non-priority area, the company will have to build another station in a priority area. However, there is no criteria or list of areas to be prioritized. By the time this report is being finalized, PL 347/2021 still needed to be approved by the Municipal Chamber. 36 of 134 C - Diagnosis 4 MOBILITY GENERAL DATA 4.1 MODAL SPLIT São Paulo’s modal distribution by main mode of travel was based on the SPMR’s 2017 Origin-Destination (OD) Survey developed by Metrô. For this analysis, the data used is from trips that had São Paulo city as an origin or as a destination, and then it was divided as: collective mode (subway, train, bus, charter transport and school bus/van); individual mode (car driver, car passenger, conventional taxi, unconventional taxi, motorcycle, motorcycle passenger and others); and non-motorized mode (cycling and walking). It is worth mentioning that the OD Survey only assesses the main mode of travel (Table 6 and Figure 11), in other words, routes that use more than one means of transport are classified according to a pre-established hierarchy9. T ABLE 6. Q UANTITY OF TRIPS ORIGINATED IN /ATTRACTED TO SÃO PAULO CITY BY MAIN MODE OF TRAVEL Modal Split Main Mode of Travel Trips in São Paulo Total Subway 3,381,585 Train 985,714 Collective Bus 5,819,414 11,452,768 Charter Transport 97,784 School Bus/Van 1,168,271 Driving a Car 5,096,941 Car Passenger 2,134,947 Conventional Taxi 83,444 Individual Unconventional Taxi 276,893 8,298,835 Driving a Motorcycle 579,858 Motorcycle Passenger 47,569 Others 79,183 Cycling 219,236 Non-motorized 8,005,894 Walking 7,786,658 Total 27,757,497 27,757,497 Source: OD Survey – Metrô (2017). 9 According to the 2017 OD Survey methodology, the hierarchy for transportation modes used in the same trip is: (1) subway; (2) train; (3) buses; (4) charter transport; (5) school bus/van; (6) taxi; (7) driving a car; (8) car passenger; (9) motorcycle; (10) bicycle; (11) others; (12) walking. Considering this hierarchy, a trip made by bus and subway will have the subway as the main mode of travel. 37 of 134 C - Diagnosis F IGURE 11. M ODAL SPLIT IN SÃO PAULO CITY Non-motorized 29% Collective 41% Individual 30% Source: OD Survey – Metrô (2017). From this division, it is possible to observe a predominance of public transport use (41.3% of total) in trips that have the city of São Paulo as an origin or destination, followed by non-motorized means of transport, such as cycling (approximately 1% of the total) and walking (28.1% of the total). Even if cycling represents only 1% of trips originated in São Paulo, there was a 45.2% increase in trips made by bicycle between 2007 and 2017 when considering all means of transport taken, not only the main mode [17]. Another modal that had significant growth in this period was transport by taxi, with a 326.7% increase of total number of trips [18]. This is due to the recent popularization of taxi apps in São Paulo – such as Uber –, which demanded the creation of a new category for 2017 OD Survey: the unconventional taxi. Even though São Paulo has a history of prioritizing individual motorized transport, trips by private car (driver or passenger) represent only a quarter of the city total (7,231,888 trips or 26.1% of total). Also, when analyzing 2007 and 2017 numbers for the city of São Paulo, there is a rising trend in public transport use, with a 14.9% increase of collective mode trips when compared to a growth of 5.9% for individual mode [17]. Part of this growth was due to the expansion of São Paulo's subway and train network in the last decade, exemplified by Metrô’s 4-Yellow Line implementation, and the extension of Metrô’s 2-Green Line and CPTM’s 9-Emerald Line. At the same time, a recent survey carried out by IBOPE Inteligência and Rede Nossa São Paulo revealed that 69% of respondents would abandon their car if there was a good public transport alternative [19] – 39% definitely would abandon their car, and 30% probably would. This percentage was even higher in 2019, before the Covid-19 global pandemic, and represented 78% of the respondents who owned private vehicles [20]. 4.2 VEHICLE FLEET For fleet data, DENATRAN (National Traffic Department) provides information about vehicles registered in a national database system called RENAVAM (National Register of Motorized Vehicles). According to DENATRAN (Table 7 and Figure 12), São Paulo’s estimated fleet is 8.8 million vehicles, of which 68% are 38 of 134 C - Diagnosis cars. Considering motorcycles, utility vehicles10 and cars, the municipality has 6.9 vehicles for every 10 inhabitants. These numbers contrast with the real demand of São Paulo residents, who mostly depend on public transport and non-motorized travel modes (Table 6 and Figure 11). T ABLE 7. S ÃO PAULO ’S VEHICLE FLEET , QUANTITY PER TYPE Vehicle Type Quantity Car 5,978,740 Truck 171,863 Utility vehicle 1,211,129 Motorcycle 1,255,867 Bus 90,520 Others 110,656 Total 8,818,775 Source: Fleet by City and Type (April 2021) – DENATRAN. F IGURE 12. S ÃO PAULO ’S VEHICLE FLEET PER TYPE BUS OTHERS 1% 1% MOTORCYCLE 14% UTILITY VEHICLE 14% CAR 68% TRUCK 2% Source: Fleet by City and Type (April 2021) – DENATRAN. Also, there is a socioeconomic dimension to the fleet distribution that must be considered, since car ownership is directly related to family income. In São Paulo, 78% of families who earn more than 5 minimum wages own a private car (Table 8), but this percentage drops to 30% when family income is less or equal to 2 minimum wages [19]. 10 For the ‘Utility Vehicle’ category it was considered a sum of caminhonetes (vehicle for cargo transport up to 3.500 kg), camionetas (vehicle for passenger and cargo transport in the same compartment), and utility vehicles (vehicle for passenger and cargo transport characterized by its versatility, even off-road). 39 of 134 C - Diagnosis T ABLE 8. P ERCENTAGE OF FAMILIES THAT OWN A PRIVATE CAR , DIVIDED PER FAMILY INCOME Owns a Family Income (in minimum wages) Total private car? More than 5 More than 2 up to 5 Up to 2 Yes 46% 78% 62% 30% No 54% 22% 38% 70% Source: Opinion poll “Viver em São Paulo: Mobilidade urbana” (2020) – IBOPE Inteligência. In addition, the 2017 OD Survey shows that people with lower income tend to use more public transport and less individual modes of travel. The summary report from Metrô [21] analyzes the modal split per income information, dividing the SPMR population into 5 groups: (1) up to 2 minimum wages; (2) from 2 up to 4 minimum wages; (3) from 4 up to 8 minimum wages; (4) from 8 up to 12 minimum wages; and (5) more than 12 minimum wages. Considering the SPMR, 72.7% of all motorized trips from group 1 were made by public transport and only 27.3% were by individual mode. For group 5, on the other hand, 75.8% of all motorized trips were made by individual means of transport. However, comparative data from 2007 and 2017 shows that this situation is slowly changing. While the richest are no longer using cars and opting for collective and non-motorized modes of travel, an overview of the SPMR shows that individual means of transport have grown among the poorest [21]. It is possible that this increase is related to tax measures to facilitate credit and reduce taxes on automobiles, which expanded access to private cars for low-income strata. Moreover, it should be noted that trips by motorcycle increased 48% in SPMR between 2007 and 2017, with a high participation of families with income of up to four minimum wages (57% of increase in this category). In the city of São Paulo, another reason for the adoption of individual means of transport among low- income families would be the high-capacity public transport network coverage (subway and metropolitan train). Based on the PNT (People Near Transit) indicator – which measures the percentage of a population that lives near public transport – measured by ITDP and WRI Brasil, only 25% of São Paulo’s residents can access medium- or high-capacity public transport by foot. Of those who have live near public transport, only 18% are from the lowest income group (from 0 to ½ minimum wage per person) [22]. Vehicles in circulation 11 in São Paulo have an average age of 12,11 years, according to data from DENATRAN (National Traffic Department) [23]. Considering the municipality’s total fleet (Table 9 and Figure 13), 14,26% of the existing vehicles are up to 5 years old, 39,61% are between 6 and 15 years old, 22,16% are between 16 and 25 years old and 23,48% are more than 26 years old. Age is an important indicator to evaluate the vehicular emission rate of an existing fleet and its environmental impact. CETESB (Environmental Company of São Paulo State) releases a yearly report of Vehicular Emissions in the State of São Paulo and it highlights that older vehicles have a higher level of air pollution emissions due to equipment deterioration and less restrictive regulations (at the time it was manufactured) [24]. In the city of São Paulo, individual cars are responsible for the largest share of GHG (greenhouse gases) emissions – 72.6% of total emissions per day, or approximately 7,253 kt of CO2e per day [25]. Buses represent 24.3% of total emissions per day, and motorcycles are responsible for the remaining 3.1%. 11 For the calculation of vehicles in circulation, vehicles with 25 years or less were considered. 40 of 134 C - Diagnosis T ABLE 9. N UMBER OF VEHICLES DIVIDED BY YEAR OF MODEL LAUNCH , CITY OF SÃO PAULO Model Year Age (in years) Nº of Vehicles % 2022 0 8,344 0,09% 2021 1 144,606 1,64% 2020 2 259,152 2,94% 2019 3 319,974 3,63% 2018 4 309,424 3,51% 2017 5 215,745 2,45% 2016 6 243,254 2,76% 2015 7 327,550 3,71% 2014 8 369,073 4,19% 2013 9 383,108 4,34% 2012 10 405,011 4,59% 2011 11 400,923 4,55% 2010 12 371,097 4,21% 2009 13 281,818 3,20% 2008 14 426,596 4,84% 2007 15 284,897 3,23% 2006 16 213,991 2,43% 2005 17 209,177 2,37% 2004 18 168,482 1,91% 2003 19 172,674 1,96% 2002 20 158,158 1,79% 2001 21 220,057 2,50% 2000 22 187,020 2,12% 1999 23 174,327 1,98% 1998 24 204,960 2,32% 1997 25 245,313 2,78% Up to 1996 More than 26 2,070,815 23,48% No information 43,223 0,49% Total 8,818,769 100% Source: Quantity of vehicles per state municipality year of model manufacturing (April 2021) – DENATRAN. 41 of 134 C - Diagnosis F IGURE 13. Q UANTITY OF VEHICLES DIVIDED BY YEAR OF MODEL LAUNCH , CITY OF SÃO PAULO 450 Thousands 400 Number of Vehicles 350 300 250 200 150 100 50 0 2022 2013 2004 2021 2020 2019 2018 2017 2016 2015 2014 2012 2011 2010 2009 2008 2007 2006 2005 2003 2002 2001 2000 1999 1998 1997 Model Year Source: Quantity of vehicles per state municipality year of model manufacturing (April 2021) – DENATRAN. 4.3 ROAD CRASHES CET annually carries out a report of traffic crashes [26]. The report is structured by the CET's Traffic Accident System (SAT), with information obtained from police reports drawn up at the Civil Police Stations and also from the records of deaths by road crashes of the IML's (Scientific Police) of São Paulo. These data make it possible to identify the location of the crash, type, vehicles involved, victims and date. With this information, CET seeks to establish priority points of intervention, target audience of campaigns, inspection locations and times, and so on. In 2019, SMT joined the Secretary of Municipal Health (SMS) to promote joint actions focused on reducing road crashes. The agreement provides for the unification of data methodologies for traffic crashes and deaths, to deepen possible diagnoses and thus develop more efficient safety policies. According to the report, in 2019 there were 758 fatal crashes in São Paulo, of which 417 (55%) were crashes with victims in vehicles and 341 (45%) were pedestrian-vehicle crashes (running over). When dividing the 791 fatal victims of these crashes by user type, the majority were pedestrians (359 or 45%), followed by motorcyclists (297 or 38%), car drivers/passengers (104 or 13%), and cyclists (31 or 4% of total). Historically, the number of traffic crashes has gradually decreased. Considering the numbers between 2010 and 2019 (Figure 14) there was a drop of 42% in the total number of fatal crashes (from 1306 to 758 annual occurrences). Regarding the total number of deaths, there was also a reduction of 42% between these years (from 1357 to 791 fatal victims). However, the total number of pedestrian deaths has risen since 2017. 42 of 134 C - Diagnosis F IGURE 14. A NNUAL EVOLUTION OF FATAL CRASHES BY TYPE 1400 1200 Quantity of Fatal Crashes 1000 800 600 400 200 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Running Over Crashes Hits Others Source: Traffic Accidents: Annual Report (2019) – CET. Two other important factors are deaths by age group, and deaths by type of user and gender. The data shows that deaths of motorcyclists are concentrated in the age group between 20 and 24 years old, in which there were 79 deaths in 2019. Most pedestrian deaths are in the age group of 40 years old and above. In relation to gender, 19% (151) of the total fatalities were female and 81% (640) were male. Regarding male fatal victims, 258 were pedestrians and 273 motorcyclists. Crashes with victims, that is, crashes that cause injuries or deaths, totalize 13,966 in 2019, 6% more than 2018 and 3% more than 2017. Of these crashes, 9,059 (58%) – more than half of the victims – were driving a motorcycle or were passengers. Regarding age and gender, most victims have between 20 and 39 years old and are male (74%). According to the report, traffic crashes are concentrated in the morning, lunch and afternoon peak hours – or from 7:00 am to 8:00 am, from 12:00 pm to 3:00 pm, and from 5:00 pm to 8:00 pm –, with the majority of crashes occurring on Friday (information for total crashes with victims, fatal or not). Among the vehicles involved, cars and motorcycles correspond to 70.6% of all fatal crashes, with 35.1% of fatal crashes involving cars (Figure 15). Of the crashes with victims, cars and motorcycles correspond to 89.8% of the total, with 51.4% involving cars. For pedestrian-vehicles crashes (running over), 58.4% of the vehicles involved are cars. 43 of 134 C - Diagnosis F IGURE 15. V EHICLES INVOLVED IN FATAL CRASHES Bike No information 4% 5% Truck 9% Car 35% Bus 11% Motocycle 36% Source: Traffic Accidents: Annual Report (2019) – CET. A spatial analysis of the topic verified that, from the 40,230 traffic crashes registered between 2017 and 2019, only 18% of crashes occurred within a 25 meters radius from a traffic light intersection (or 7,052 crashes out of 40,230). This means that 2,711 intersections in São Paulo (or 46% of 5,886 total intersections) registered some type of crash. These values were extracted from geoprocessing studies12 and are shown in Table 10. T ABLE 10. RELATION BETWEEN CRASHES AND TRAFFIC LIGHT INTERSECTIONS (CITY OF SÃO PAULO) Intersections Nº of crashes % of crashes with crashes % of total within 25 Total crashes within 25 Year within 25 intersections meters from an per year meters from meters intersection an intersection 2017 2,021 34% 3,440 13,271 26% 2018 1,352 23% 1,834 13,046 14% 2019 1,291 22% 1,778 13,913 13% Total - - 7,052 40,230 18% Source: CET. 12 This analysis was done with geoprocessing tools and the traffic light geodatabase elaborated for this study. For more details about the sources of this geodatabase, see item 5 of this report. 44 of 134 C - Diagnosis 5 TRAFFIC LIGHTS DATA COLLECTION The city of São Paulo owns a large traffic light network, composed of over 6,500 signalized intersections and thousands of traffic controllers. Given that São Paulo has a population of 12 million habitants that travel daily by multiple means of transportation (car, bus, bicycle, walking and so on), the traffic light system is of great importance for the entire city operation. In this context, studies that seek adequate solutions for São Paulo are vital, as new policies focused on evolving the technological capability of this system are being developed. To build a coherent analysis of the traffic light system current situation, a number of sources were consulted and cross-referenced. This chapter portrays documents that were key references to the consolidation of an updated and detailed traffic light geodatabase, providing information such as location, infrastructure, failures, management, and so on. In addition to traffic light reports – such as IDOM study, developed in partnership with World Bank –, public biddings, official online platforms, and private-public partnership studies were consulted. For a better understanding of the chronology, Figure 16 presents some important dates of documents analyzed. F IGURE 16. C HRONOLOGY OF DATA SOURCES Source: Future Mobility for São Paulo Consortium. 5.1 TRAFFIC LIGHTS PPMI Firstly, São Paulo City Hall has two instruments for the presentation of studies made by private companies: the Procedure for Expression of Interest (PMI) and the Preliminary Procedure for Expression of Interest (PPMI). These two instruments, regulated by Decree nº 57.678/2017, serve to assist the public administration in structuring privatization projects, concessions, and Public-Private Partnerships. The PMI and PPMI studies do not guarantee any kind of advantage for these companies in future biddings or tenders. T ABLE 11. T YPES OF EXPRESSION OF INTEREST (PMI AND PPMI) PMI PPMI ▪ Studies can be carried out without prior ▪ Studies can only be carried out with the approval of São Paulo City Hall. approval of São Paulo City Hall. ▪ The PPMI is convened when it could be useful to obtain specific preliminary subsidies. 45 of 134 C - Diagnosis PMI PPMI ▪ A PMI is convened to obtain in-depth ▪ The studies will be published without subsidies in specific subjects, or to enable an prejudice to intellectual property. integrated structuring. ▪ Studies used in the project modelling will not ▪ The studies will be published as soon as the be remunerated. final public notice of the privatization project is launched for bidding. ▪ Studies used in the project modelling will be paid by the company who wins the privatization contract. Source: SMDP, https://www.prefeitura.sp.gov.br/cidade/secretarias/governo/desestatizacao_projetos/. In February 2018, a public call for PPMI was published to gather preliminary subsidies for the structuring of a private partnership aimed at the “modernization, maintenance and availability of the traffic light network” in São Paulo. According to SP Parcerias – company associated with SGM in charge of structuring and developing Public-Private Partnerships, privatization, and concession projects –, the PPMI was organized due to the current state of São Paulo’s traf fic light system, currently with no real-time automation13, outdated and weather sensitive. Before the study submission, SMDP – responsible for organizing the PPMI – received a request for additional documents and information about: ▪ (1) Traffic light maintenance contracts in force. ▪ (2) List of existing equipment, infrastructure and logistic centers related to the traffic light network (operational centers, controllers, cameras, optical fiber ducts, metallic cabling, etc); and those that could be integrated due to synergies (tunnel signaling equipment, tunnel monitoring centers, etc). Answering to this request, CET informed that the traffic light system maintenance was established by contracts nº 062/17, nº 063/17 and nº 064/17, each contract responsible for one geographical area of the city as determined by Bidding nº 023/17 – “Provision of continuous services, traffic light system equipment and infrastructure maintenance with supply of materials, in the city of São Paulo”. At the time, digital documents were also included to answer request (2), among them: ▪ Traffic light registration: MS-Excel spreadsheet containing all signalized intersections in the city of São Paulo and information on: intersection’s address; geographic coordinates; Traffic Engineeri ng Management (GET); Traffic Light Control Department (DCS); controller model; possibility of centralization; quantity of vehicular and pedestrian signal heads; quantity of pushbuttons; among other information related to traffic light infrastructure. ▪ Control and monitoring centrals: List of existing traffic light control and monitoring centrals and their addresses. ▪ CCTV registration: List of CCTV cameras’ location, type – analog or digital – and indication of which control and monitoring central they are connected to. ▪ Registered duct network: Autocad file with the duct network map. 13 Even though SP Parcerias [27] and IDOM [28] both state that São Paulo currently does not work with real-time traffic control, the city had a functioning real-time traffic control network that was implemented by the end of 1990 – a modernization program called ‘Projeto CTA’ [31]. 46 of 134 C - Diagnosis The basis for our database was the traffic light registration, from which it was possible to obtain information about the signalized intersections’ identification, location, controllers, and physical infrastructure. Later, the database was georeferenced as points – each point corresponding to an intersection – according to their geographic coordinates (latitude and longitude). The resulting geodatabase was compared to the traffic lights shapefile provided by Geosampa for verification and, despite minor differences, the overall point arrangement matched. Since Geosampa’s shapefile was from January 2017 and had less information, the traffic light registration was chosen as the main reference for this study. It is important to stress that both shapefiles – the geodatabase of this study and the one provided by Geosampa – presented minor errors in point placement that could affect a small fraction of geographical analysis results. However, it was assumed that these displacements were due to registration failures (in geographic coordinates, for example) and could not be corrected manually. Since the nature of this study is to evaluate the traffic light system of São Paulo in a macro scale, these minor errors were accepted as part of the geodatabase. In June 2018, eight companies had submitted studies for the Traffic Lights PPMI: Cisco Siemens, Seebot, Serttel, Zeev Consult Engenharia, Egis Engenharia e Consultoria, Engie (in partnership with FiscalTech) and SPIn Soluções Públicas Inteligentes (in partnership with Kapsch, TrafficCom, SITRAN and Vallya Advisors). As a result, SMDP elaborated an evaluation report in which all subsidies were compared according to their converging and diverging points. SPIn results were not published due to the in-depth content of their study, similar to a PMI. According to SP Parcerias’ Annual Governance Letter from 2018 [27], after the Traffic Light PPMI was concluded, SMT was not interested in continuing the privatization proposal since there was no consensus about the best technology to be implemented. In the same document, a World Bank study proposal is mentioned, a pilot project focused on traffic light technology funded by the UK Prosperity Fund in the context of the Future Cities Project. 5.2 CONSULTANCY SERVICES FOR ALTERNATIVES ANALYSIS AND OTHER STUDIES FOR TRAFFIC LIGHTS SYSTEMS IN SÃO PAULO (IDOM STUDY) The IDOM study is a series of reports developed by Spanish consulting company IDOM for the analysis and identification of possible technical solutions to be employed in the city. This study was commissioned by World Bank and developed between 2018 and 2019 alongside CET and SMT. São Paulo City Hall had much interest in this study, since the Traffic Lights PPMI subsidies didn’t reach a general agreement regarding technical solution. The IDOM study is composed by: ▪ D1 – Status Quo Report ▪ D2 – Analysis of Technological Alternatives ▪ D3 – Economic Preliminary Analysis ▪ D4 – Methodology for Evaluation of Performance and Performance of Adaptive Systems of Semaphore Control in Real Time ▪ Final Executive Summary For this particular stage, the “D1 – Status Quo Report” [28] will be referenced as it contains relevant information regarding CET’s operation and primary data for the traffic light system. The report, released in February 2019, consists of: 47 of 134 C - Diagnosis ▪ Description of the city: Presentation of the main characteristics of the city of São Paulo, geographically, economically and of mobility, in order to refer the city to other urban realities that could be of reference. ▪ Mobility Agents: Collection of identification of agents involved in mobility management. ▪ Normative framework: This section identifies the norms and laws that are applicable in the traffic light system and, therefore, presupposes criteria for the selection of technological alternatives. ▪ Traffic light system: Detailed presentation of the characteristics of the traffic light system and other infrastructures related to its implementation, such as the communication network and the power supply. After report D1, the IDOM study focused on selecting technological solutions for traffic light systems that would suit São Paulo city. Their conclusions were: ▪ Centralized traffic light system, with optical fiber connection for controllers and cameras. ▪ Define a standard communication protocol – UTMC2 has an important advantage as 900 controllers in the city are already adapted to it. ▪ Study adaptive mode alternatives and implement some particularization of SCOOT for the city. ▪ Install other types of sensors (virtual sensor or radiofrequency sensors). ▪ Study implementation of public transport priority at intersections. The technological alternatives study was followed by an economic analysis and a benefit evaluation. Its main conclusion was that SCOOT system is the most beneficial solution for São Paulo due to CET’s technical requirements, supplier availability, and UTMC2 protocol compatibility. Moreover, IDOM stressed the importance of preventive and corrective maintenance of São Paulo’s traffic light system. Therefore, as a recommendation, city officials should create a fund for annual investments in maintenance. The IDOM study was developed in the same context as this report, as a partnership between World Bank and the UK Government Prosperity Fund. Thus, results from IDOM study are being integrated into other documents of this study, such as report B – Benchmarking. 5.3 MODERNIZATION AND MAINTENANCE TENDER In October 2019, a new bidding for the modernization and maintenance of traffic light control system in São Paulo was published by CET due to the need for revitalization and technological upgrade of the intelligent system infrastructure. The Bidding nº 01/19 geographically divided the city in four lots (Annex C3. Modernization and Maintenance Tender Map and Figure 17), each company responsible for the traffic light system maintenance and modernization inside the corresponding area. The winning company would be decided according to lowest global price offered per lot, but CET’s estimated value is approximately BRL 936 million. As an annex to Bidding nº 01/19, the Terms of Reference (ToR) contains updated information for the traffic light system Status Quo analysis and provides a comprehensive picture of CET’s requirements for the system modernization. Firstly, the ToR mentions the last traffic light system modernization contract that renovated 4,800 intersections from 2013 to 2015, and the resulting operation improvement – up to 70% of failure reduction in 2015 when compared to previous numbers. However, the lack of resources for further modernization and maintenance investments resulted in an exponential increase of traffic light failures, consolidating a vulnerable point in São Paulo’s traffic flow. In 48 of 134 C - Diagnosis this context, the Tender establishes a strategy to prioritize areas with a higher concentration of failures: the expanded city center of São Paulo, limited by the Mini Ring Road. According to the ToR, São Paulo currently has 5,895 intersections with regular traffic lights and 603 flashing yellow traffic lights for warning. Within the Mini Ring Road – formed by Marginal Pinheiros, Marginal Tietê, Bandeirantes Avenue and others – there are 2,583 intersections that should be modernized to reduce failures and allow new control strategies according to the region (Annex C3. Modernization and Maintenance Tender Map). From the Traffic Lights PPMI registration table – foundation of the current working database –, 5,886 intersections with regular traffic lights and 603 flashing yellow traffic lights were identified, in other words, there is a difference of 9 intersections between 2018 and 2019. While it is not possible to pinpoint the modifications carried out during this period, the total value is close enough to validate the working database. Another source of information is the detailed list of intersections to be modernized – ‘Attachment B’ from ToR – that contains: local ID; address; Traffic Engineering Department (DET); Traffic Light Control Department (DCS); and proposed control type. This list was cross-referenced with the working database to identify all intersections to be renovated. According to this analysis, there are 2,571 intersections contemplated by the Modernization Tender inside São Paulo’s Mini Ring Road – 12 less than what was mentioned in the ToR (Annex C3. Modernization and Maintenance Tender Map). Each modernized intersection must follow the proposed control type, which are: ▪ Real time system (TR): Intersections must be communicating with the operational central with SCOOT system fully loaded. ▪ Centralized fixed time system (TFC): Intersections must be working centralized, allowing remote time adjustments and failure corrections. ▪ Monitored fixed time (TFM): The GPRS module must send online information of the equipment status to an operational central. The modernization scope also covers: ▪ Complementation and adequation of the communication and connectivity network, connecting all traffic centrals and the entire traffic light network. For this, all optical fiber sections within the Mini Ring Road must be constructed underground and according to the service provision progression. ▪ Modernization of the current traffic light system located inside the Mini Ring Road. This includes the expansion of intersections with UPS system and replacement of old controllers for technological upgrade. ▪ Revitalization of vehicle detection systems to meet the requirements of an Adaptive System in Real Time. ▪ Software update for Adaptive System in Real Time technology in traffic control centrals. The standard adopted in São Paulo is SCOOT, which allows a continuous and permanent adaptation of traffic light times and the introduction of specific parameters for bus prioritization in signalized intersections. ▪ Adjustment of traffic light central’s equipment to ensure the centralization and monitoring of new controllers. CET carried out a study to identify which modifications are necessary for each intersection, so not all intersections listed will receive a full modernization. 49 of 134 C - Diagnosis F IGURE 17. M ODERNIZATION AND MAINTENANCE TENDER MAP Source: Future Mobility for São Paulo. 50 of 134 C - Diagnosis In addition to the modernization work, the Tender also includes maintenance services, which are responsible for keeping the traffic light system without failures 99% of all times. To reach this level of performance, the maintenance scope covers repairs, substitution, and cleaning of: ▪ Fixed time controllers with or without GPRS (all components). ▪ Adaptive controllers (all components). ▪ Signal heads and its complements. ▪ GPRS modules, tests, connectivity repair with operators and maintenance central. ▪ Poles, totems, cantilever arms with all necessary infrastructure. ▪ Duct access boxes. ▪ Underground duct system for electrical and communication. ▪ Underground or aerial cabling for electrical, communication (metallic, optical fiber), internal wiring, connectors, terminals, and circuit breakers. ▪ Push buttons for pedestrians, simple and sound modules. ▪ Other accessories that complement the traffic light system and its integrated systems. Also, the contracted company will be responsible for the electrical maintenance of other equipment, such as: traffic controllers’ UPS systems, flashing yellow traffic lights for warning, vehicular detection system, cameras, among others. This service must be provided 24 hours per day, 7 days per week in all signalized intersections for the contract’s duration – 60 months. Basically, it presents the same structure for current maintenance contracts in São Paulo. According to Bidding nº 01/2019, interested companies had until January 16, 2020 to submit their proposals. However, two days prior to this limit, CET suspended the Tender indefinitely for reanalysis and adjustments. 5.4 SMART CITY PMI In October 2020, SGM published an open call for a Smart City PMI to gather operational and financial studies for the requalification, operation, and maintenance of São Paulo’s duct infrastructure; implementation, operation, and maintenance of the optical fiber network for data transmission to PMSP access points; and commercial exploitation of both optical fiber network and duct infrastructure. For this, companies would need to submit four different studies: ▪ Engineering modelling: Preliminary engineering study encompassing the duct infrastructure requalification; optical fiber network implementation to connect PMSP access points (public administrative buildings); and, if necessary, an expansion plan for the underground duct infrastructure. In addition, each company must present studies for the implementation of optical fiber between CET’s traffic light equipment and their respective operation centrals – only applicable to equipment located within 500 meters from the underground duct infrastructure. ▪ Operational modelling: This study must present procedures and relevant information for the operation and maintenance of the duct infrastructure; optical fiber network for PMSP access points; and optical fiber network for traffic light and monitoring equipment. Also, it may include new public or commercial uses for optical fiber, such as pollution monitoring, public Wi-Fi hotspots expansion, solid waste tracking, among other uses. ▪ Subsidies for economic-financial modelling: This study must introduce necessary subsidies for an economic-financial modelling of the privatization project, including a market potential diagnosis and subsidies for a complete viability analysis. In addition, it should provide business models for the duct infrastructure and optical fiber network commercial exploitation. 51 of 134 C - Diagnosis ▪ Subsidies for legal modelling: Diagnosis report of legal-regulatory scenario and subsidies for legal modelling. To support these PMI studies, some documents were published by SGM, such as: ▪ PMSP access points: List of public buildings and equipment to be connected with optical fiber. ▪ SPTrans: Map in Google Earth file of SPTrans duct infrastructure. ▪ CET ducts and boxes: Map in Google Earth file of CET duct infrastructure. Since communication network for traffic lights is a relevant topic for this study, the CET ducts and boxes mapping was incorporated for further analysis as it contains information about connection between traffic light equipment and operational centrals (Annex C11. Duct Network Map). However, the Terms of Reference specifies that each engineering modelling study must include an updated duct infrastructure registration (obtained by field inspection), indicating that the provided mapping does not reflect the current status of operation. Another important topic is that, according to information provided by this PMI, São Paulo’s public duct infrastructure is limited to SPTrans and CET’s communication network (bus and traffic operation, respectively). In June 2021, it was announced that a total of eight companies were authorized to submit studies for Smart City PMI: CCBR Catel Construções do Brasil SA. and BPMI Infra S.A.; Control Home Ambientes Inteligentes LTDA.; Lume Consultoria Eireli ME; ME Medaglia Serviços de Engenharia de Sistemas LTDA.; Moyses & Pires Sociedade de Advogados; Poyry Tecnologia LTDA.; Radar PPP LTDA.; and SPIn Soluções Públicas Inteligentes Consultoria LTDA. By October 2021 all studies must be submitted, but they will only be published after the consolidation of a future privatization bidding. 52 of 134 C - Diagnosis 6 TRAFFIC LIGHTS CONTRACTS Considering all necessary services to maintain an operating traffic light system in São Paulo, SMT and CET are responsible for contracting companies to provide the maintenance, supply, and implementation of various systems. For the past decade, most of these services have been arranged in a single maintenance contract geographically divided by lots. Each winning company or consortium provides the contracted services within a delimited area of São Paulo – respective lot. However, the last three biddings for maintenance contracts have had differences in service scope, duration, and bidding process. Regarding service scope, these maintenance contracts can include a modernization agenda depending on the available public resources. Even though corrective maintenance is always included in these contracts, the modernization scope – items to be updated and coverage of renovation – may vary. For instance, the Traffic Lights Renovation from 2013-2015 covered 4,800 intersections, but focused mainly on electrical systems (electrical wiring, UPS), GPRS communication modules, implementation of 880 fixed time controllers, 120 real time controllers, and software provision for the maintenance central. On the other hand, the Modernization and Maintenance Tender from 2019 (suspended) covers the modernization of traffic light intersections located in São Paulo’s expanded city center while focusing on underground duct infrastructure (civil works for communication and electrical ducts), expansion of optical fiber network, substitution of 1,020 controllers (610 real time and 410 fixed time), implementation of detectors (inductive and virtual loops), and real time control system (SCOOT). Currently, there is no modernization contract in place, only maintenance. All contracts mentioned include supply of materials, labour and civil works needed. As for bidding process, SMT or CET will open a public call for companies and publish the Terms of Reference – document that contains all technical information about services and materials to be provided, it will be the basis for price calculation. On a predetermined date, all price proposals will be analyzed and compared (in a digital platform or in person). Depending on the dispute mode, price proposals can be lowered through successive bids. For maintenance contracts, winning companies are decided based on smallest global value per lot. However, these bidding processes have not been consistent considering the last three maintenance contracts, especially regarding dispute mode, number of lots, execution regimen, and duration. T ABLE 12. C OMPARISON OF BIDDING PROCESSES FOR THE LAST THREE TRAFFIC LIGHT MAINTENANCE CONTRACTS Modernization and Current Maintenance Bidding item Maintenance Tender Traffic Lights Renovation Contracts (suspended) Contractor CET CET SMT Year 2019 2017 2013 Provision of continuous Provision of services, traffic light Contracting services for modernization and system equipment and traffic control signaling Description maintenance services for infrastructure system renovation, with the traffic light system in maintenance with supply supply of materials the city of São Paulo of materials, in the city of São Paulo 53 of 134 C - Diagnosis Modernization and Current Maintenance Bidding item Maintenance Tender Traffic Lights Renovation Contracts (suspended) Document Bidding nº 01/19 Bidding nº 023/17 Bidding nº 02/2013 Number of 4 3 3 lots Judgment Smallest global value per Smallest global value per Smallest global value per criteria lot lot lot Closed – bidders present “Pregão eletrônico” – “Pregão presencial” – their commercial modality where bidders modality where bidders Dispute mode proposals, no possibility of can make successive bids can make successive bids successive bids in a digital platform in person Contract for unitary price Contract for unitary price Execution – when there is no – when there is no Contract for global lot regimen precision in quantity of precision in quantity of price materials and services materials and services Duration 60 months 12 months 24 months Source: CET Bidding nº 01/19, CET Bidding nº 023/17, SMT Bidding nº 02/2013, and CET RILCC (Internal Regulation for Biddings, Contracts, and Partnerships). Nonetheless, maintenance contracts are not the only way to provide services for the traffic light systems in São Paulo. Whenever there are traffic impact analyses for future buildings (trip generation studies), countermeasures can be applied as traffic light implementation. Each case is unique, so there isn’t a standard model for these projects, but CET is always in charge of management and scope definition. CET also has a technical team that can implement new traffic light equipment or do maintenance in a smaller scale. However, when CET’s team is responsible for these services a supply contract is needed for materials. These types of contracts are not frequent and may vary according to the situation. According to the IDOM study [28], two types of communication services are also contracted by public biddings (CET): GPRS and TETRA systems. The GPRS network is used for traffic light maintenance – transmission of report failures from the controllers – and information exchange for variable message signs. There are approximately 1,200 lines that are renovated every two years. As for the TETRA service, it is reserved for operational communication of the signaling department. ▪ Maintenance Contracts in Force São Paulo currently has three contracts for traffic light system maintenance, defined by CET Bidding nº 023/17 – “Provision of continuous services, traffic light system equipment and infrastructure maintenance with supply of materials, in the city of São Paulo” – published in June 2017. Before this bidding, CET had been in charge of traffic light system maintenance since December 2016, when the last contract ended. With no replacement, CET maintenance team was not able to keep up with the demand resulting in a series of “traffic light blackouts” during the first half of 2017. The bidding geographically divided São Paulo in three lots, so each contract is responsible for the traffic light system maintenance inside the corresponding area. Maintenance services must keep the traffic light system working, as well as its components. It may include localized repair, substitution, and cleaning of:  Fixed time controllers with or without GPRS.  Adaptive controllers.  Signal heads for flashing yellow traffic lights. 54 of 134 C - Diagnosis  Signal heads and its complements.  UPS systems (whole set).  GPRS modules, tests, connectivity repair with operators and maintenance central.  Reversible warning luminous signs.  Flood warning luminous signs.  Barrier gate lighting.  Pedestrian crossing lighting.  Poles, totems, cantilever arms with all necessary infrastructure.  Duct access boxes.  Underground duct system for electrical and communication.  Underground or aerial cabling for electrical, communication (metallic, optical fiber), internal wiring, connectors, terminals, and circuit breakers.  Push buttons for pedestrians, simple and sound modules.  Other accessories that complement the traffic light system and its integrated systems. In addition, the contracted company will be responsible for the electrical maintenance of other equipment, such as: warning luminous signs, inductive loops, virtual loops, among others. The replacement or reconstitution of loop detectors are not included in the bidding. Moreover, all services and equipment provision must take into consideration current regulations and CET’s technical specifications. For each traffic light failure detected, CET will issue a Service Order – which will describe type of service and location. There are two distinct moments for each Service Order: the first call, in which there is no need for equipment substitution (except electronic modules for traffic controllers, supplied by CET); and the second call, when it is impossible to correct a failure without equipment substitution and a second visit is needed. For this, a new Service Order is issued. The services provided are classified according to priority and must be concluded within a limited time period: ▪ Priority 1: Maintenance service that could offer risk to the safety of people or goods. Execution must not exceed two hours. ▪ Priority 2: Maintenance service that, if not executed, could cause Priority 1 failures. Execution must not exceed 12 hours. ▪ Priority 3: Maintenance service that, if not executed, will compromise traffic light signaling. Execution must not exceed 24 hours. After every Service Order call, the contracted company must provide a photographic report showing the ‘before’ and ‘after’ situation for analysis. Maintenance services must also cover damage caused by external agents, such as traffic collisions, failures in high-voltage networks, and vandalism. However, in these cases, CET will reimburse all costs for service provision and equipment supply. Maintenance must be provided 24 hours per day, 7 days per week in all signalized intersections for the contract’s duration – 12 months. After the bidding process conclusion, three contracts were signed for the maintenance of São Paulo’s traffic light system. Even though the contracts’ duration was for 12 months, they have been renovat ed every year for the lack of a new maintenance bidding. 55 of 134 C - Diagnosis T ABLE 13. T RAFFIC LIGHTS MAINTENANCE CONTRACTS IN FORCE Bidding Contract Contractor Company Contracted Date Duration Value (BRL) Lot Number MCS Inteligente Consortium (Meng Engenharia Comércio e Indústria, CLD CET Lot 01 062/17 Construtora Laços 17/08/2017 12 months 10,791,710.84 Detetores e Eletrônica LTDA., and Sinalronda Sinalização Viária e Serviços LTDA.) Semáforo Paulistano Consortium (Serttel LTDA., Sitran Sinalização de Trânsito CET Lot 02 063/17 17/08/2017 12 months 10,790,534.37 Industrial LTDA., and Sigma Engenharia Indústria e Comércio LTDA.) ARC Comércio Construção e CET Lot 03 064/17 11/08/2017 12 months 10,791,192.31 Administração de Serviços LTDA. Source: Portal da Transparência, PMSP. The pricing of each contract must include all direct and indirect costs of services provided in their respective lot, such as equipment supply, technical team, transportation, labour benefits, and so on. Table 14 shows in detail the description and cost of covered services for the first call (Item 10 of the table), when there is no need for equipment replacement, and services for the second call (Items 1 to 9 of the table), when equipment substitution is necessary for the repair. T ABLE 14. P RICING OF SERVICES FOR TRAFFIC LIGHT MAINTENANCE , CONTRACTS Nº 062/17 ( LOT 01), Nº 063/17 (LOT 02) AND Nº 064/17 ( LOT 03) Item Description Lot 01 (BRL) Lot 02 (BRL) Lot 03 (BRL) 1 Supply with controller substitution 2,071,615.74 2,116,482.94 2,027,564.07 Supply with UPS substitution and 2 631,869.84 1,032,692.96 596,037.28 removal Supply with support elements 3 167,880.82 275,657.40 182,135.47 substitution Supply with complementary elements 4 49,731.36 67,841.28 54,361.44 substitution Supply with data transmission system 5 155,488.32 234,289.12 184,523.76 substitution Supply with electrical network 6 883,770.89 657,919.73 917,401.66 substitution 56 of 134 C - Diagnosis Item Description Lot 01 (BRL) Lot 02 (BRL) Lot 03 (BRL) 7 Supply with signal head substitution 847,289.79 808,533.18 864,431.43 Supply with pedestrian crossing lighting 8 213,002.48 352,163.76 298,953.28 substitution 9 Duct network substitution 638,656.00 603,200.00 598,880.00 10 First call 5,132,405.60 4,641,754.00 5,066,903.92 TOTAL 10,791,710.84 10,790,534.37 10,791,192.31 Source: Portal da Transparência, PMSP. 57 of 134 C - Diagnosis 7 TRAFFIC LIGHTS CHARACTERIZATION For the characterization of the intersections in São Paulo, different types of indicators were used in order to consider aspects of the intersections’ operation and its infrastructure. Both from the point of view of the road infrastructure through roads hierarchy and the availability of infrastructure for public transport, as well as from the point of view of the available communications infrastructure. The way in which users use the infrastructure was also considered troughout the predominant means of transport in each of the districts, based on the results of the city's origin/destination survey. In addition, the planning of the city was also considered based on the predominant uses of land in the districts. 7.1 INPUTS FOR TRAFFIC LIGHTS CHARACTERIZATION Multiple factors were taken into consideration as a basis for São Paulo’s traffic lights characterization, which will be a reference for future technological recommendations. To identify key aspects for this analysis, urban mobility and traffic control themes were researched to provide insights about traffic light operation in the city. Furthermore, the traffic light geodatabase – consolidated with data collection specified in Chapter 5 – was cross-referenced with georeferenced information published by São Paulo City Hall. As a result, this section will present diagnoses of topics that were relevant for the development of a traffic light characterization methodology. 7.1.1 Road and Traffic Light System Diagnosis 7.1.1.1 Road Hierarchy Based on the Brazilian Traffic Code (CTB) established parameters, Law nº 9503/97, São Paulo defined the city’s road classification through Ordinance DSV.G 18/19, published in February 2019. Therefore, according to Attachment I from Ordinance DSV.G 18/19, urban routes that are open to circulation are classified as: ▪ (I) Rapid Transit Road (VTR): There are no level intersections, no direct access to properties and no level pedestrian crossings. VTRs are more suitable for fluidity and long-distance connections, such as connections between highways and city regions. ▪ (II) Arterial: Arterials have level intersections, usually controlled by traffic lights, and property access. They also have access to secondary and local roads, enabling traffic between different city regions. Predominance of passing traffic, where structural corridors for public transport are located. ▪ (III) Collector: It collects and distributes traffic that has the need to enter or exit VTRs or arterial roads, enabling traffic within the city regions. It simultaneously allows vehicle flow and property access, connecting neighborhoods to the main city roads. ▪ (IV) Pedestrian: Roads or set of roads intended for the circulation of pedestrians. They have different physical characteristics or signaling that restricts vehicle flow, allocating these areas to pedestrians. ▪ (V) Local: Local roads do not have traffic lights in level intersections, as they are intended for local access or restricted areas. There is no passing traffic (preferably residential). According to data collected by CET in October 2019, the road network in São Paulo has approximately 20 thousand kilometers of roads, of which 200 km correspond to VTRs (1%), 2,500 km are Arterials (12%) and 4,000 km are Collectors (20%). Most of the road network is made up of Local roads, with more than 13 thousand kilometers in length (65%). Only a small portion is dedicated to Pedestrian roads, which add up 58 of 134 C - Diagnosis to about 30 km without counting alleys, urban stairs, etc. Distribution of road hierarchy in São Paulo can be seen in Annex C4. Road Hierarchy Map and Figure 18. In order to categorize each signalized intersection according to road hierarchy, georeferencing operations were carried out to associate the position of traffic lights (vector layer, points) with São Paulo’s road network (vector layer, polylines). For this, the Traffic Lights PPMI registration table – which holds information about every single traffic light in São Paulo – was georeferenced as points through its geographic coordinates. This geodatabase for signalized intersections will be used for further analysis. For each intersection one or more roads were identified, and the main classification was determined according to the road with higher hierarchy. For instance, if a traffic light was located at the junction of an arterial road and a collector, it would be classified as an arterial. Then, these intersections were divided into Traffic Engineering Management (GET) areas of São Paulo, as established by CET. For more information on this, see item 7.1.1.2 of this report. The result of this analysis is shown in Table 15 and Figure 19. It is visible that most traffic lights in São Paulo are implemented at arterial roads (58.2%), followed by collectors (40.7%). Together they represent 98.9% of all traffic lights in the city. T ABLE 15. Q UANTITY OF SIGNALIZED INTERSECTIONS BY ROAD HIERARCHY GET Highway Arterial Collector Local Total CN 2 423 417 10 852 NO 0 400 301 6 707 SE 8 525 373 10 916 SU 0 545 331 12 888 SO 0 519 401 7 927 MB 0 63 1 0 64 LE 0 527 272 5 804 OE 0 426 297 5 728 Total 10 3,428 2,393 55 5,886 Source: Traffic Lights PPMI (2018) and CET (2019). 59 of 134 C - Diagnosis F IGURE 18. R OAD HIERARCHY MAP Source: Future Mobility for São Paulo. 60 of 134 C - Diagnosis F IGURE 19. P ROPORTION OF SIGNALIZED INTERSECTIONS BY ROAD HIERARCHY Local Highway 1% 0% Collector 41% Arterial 58% Source: Traffic Lights PPMI (2018) and CET (2019). 7.1.1.2 Traffic Light Network Traffic Engineering Management (GET) The Traffic Engineering Company (CET) is a Municipal Government indirect administration entity responsible for planning and implementing traffic operation in São Paulo’s road network, ensuring traffic safety and fluidity. For this, CET’s Traffic Engineering Superintendence – that belongs to the Operations Directory – is divided into Traffic Engineering Management units (GETs) that carry out traffic operation field activities in a determined area of the city. Over the last century, São Paulo urban growth was based on urban policies that prioritized motorized modes of transport, both individual and collective. In this context, CET was created in the 70’s focusing its operation on traffic management and on the perception of traffic as an imminent and public problem. However, after some strategic changes in 2013, CET underwent an operational restructuring built on prioritization of public transport, pedestrians, and cyclists. Thus, instead of basing its activities on traffic operation, CET starts to manage the road system under the logic of traffic corridors’ – high-capacity roads that concentrate long/medium-distance trips and are often public transport axis [29]. Among the main changes caused by this new strategic guideline is the spa tial reorganization of GET’s management areas, aiming to create geographical regions that encompassed traffic corridors in their entirety and preserving corridor continuity in São Paulo’s radial road structure. In this context, here follows the description of each GET: ▪ GET-CN: Covers the center of São Paulo (Republic, Sé and Bom Retiro districts) and part of the northern area (Casa Verde, Santana, Cachoeirinha, Mandaqui, Tucuruvi, Vila Maria and so on). It includes Cruzeiro do Sul Avenue, Santos Dumont Avenue, São João Avenue, Bandeiras Bridge, and Governador Orestes Quércia Bridge. It does not include Marginal Tietê, Presidente Dutra Highway, Fernão Dias Highway, and 9 de Julho Avenue. ▪ GET-NO: Covers the northwest region of São Paulo (Perdizes, Barra Funda, Vila Leopoldina, Pirituba, Freguesia do Ó, Jaraguá, Perus and so on). It includes the Pirituba/Lapa/Centro bus corridor, Inajar/Rio Branco/Centro bus corridor, Remédios Bridge, and Júlio de Mesquita Neto Bridge. It does not include Marginal Tietê, Bandeirantes Highway, Anhanguera Highway, and São João Avenue. 61 of 134 C - Diagnosis ▪ GET-SE: Covers the southwest region of São Paulo (Mooca, Ipiranga, Vila Prudente, Carrão, Sapopemba, Aricanduva and so on). It includes the Paes de Barros bus corridor, the northern section of Tiradentes Express, Aricanduva Avenue, Itaquera Avenue, Bresser Viaduct, and Guadalajara Viaduct. It does not include Antônio Abdo Viaduct, Salim Farah Maluf Avenue, Presidente Tancredo Neves Avenue, José Alencar Gomes da Silva Avenue, and sections of Professor Luiz Ignácio Anhaia Mello Avenue/BR-050 that form the Mini Ring Road. ▪ GET-SU: Covers the southern region of São Paulo (Liberdade, Vila Mariana, Saúde, Jabaquara, Campo Grande, Cidade Dutra, Grajaú and so on). It includes the Rio Bonito bus corridor, Washington Luís Avenue, 23 de Maio Avenue, República do Líbano Avenue, and Indianópolis Avenue. It does not include Marginal Pinheiros, Presidente Tancredo Neves Avenue, Afonso D’Escragnolle Taunay Avenue, Bandeirantes Avenue, Ibirapuera Avenue, Imigrantes Highway, and the Ayrton Senna Road Complex. ▪ GET-SO: Covers the southwest region of São Paulo (Itaim Bibi, Moema, Santo Amaro, Jardim São Luís, Jardim Angela and so on). It includes the Santo Amaro/9 de Julho bus corridor, Ibirapuera bus corridor, Berrini bus corridor, Ponte Baixa bus corridor, Guarapiranga bus corridor, Giovanni Gronchi Avenue, Carlos Caldeira Filho Avenue, Itapecerica Road, Morumbi Bridge, João Dias Bridge, and Socorro Bridge. It does not include Marginal Pinheiros, Bandeirantes Avenue, and Paulista Avenue. ▪ GET-MB: Covers Marginal Tietê, Marginal Pinheiros, Bandeirantes Avenue, and other roads that constitute São Paulo’s Mini Ring Road. ▪ GET-LE: Covers the east region of São Paulo (Brás, Belém, Penha, Itaquera, São Miguel, Cidade Tiradentes and so on). It includes Jacu-Pêssego Avenue, José Alencar Gomes da Silva Avenue, Alcântara Machado Avenue, and Radial Leste Avenue. It does not include the Ayrton Senna Highway, Engenheiro Alberto Badra Viaduct, Salim Farah Maluf Avenue, Guadalajara Viaduct and Bresser Viaduct. ▪ GET-OE: Covers the western region of São Paulo (Pinheiros, Butantã, Jaguaré, Rio Pequeno, Vila Sônia, Campo Limpo and soo n). It includes the Rebouças/Campo Limpo/Centro bus corridor, Doutor Arnaldo Avenue, Heitor Penteado Street, Queiroz Filho Avenue, Paulista Avenue, Jaguaré Bridge, and Eusébio Matoso Bridge. It does not include Marginal Pinheiros and Raposo Tavares Highway. GETs are divided into Traffic Engineering Departments (DETs) that are operational subunits of each management area (GETs). In total, São Paulo is composed of 27 DETs distributed in 8 GETs (Annex C5. Management Areas Map and Figure 20). T ABLE 16. I DENTIFICATION OF TRAFFIC E NGINEERING MANAGEMENT UNITS (GETS) AND THEIR RESPECTIVE TRAFFIC E NGINEERING D EPARTMENTS (DETS ) GET DET Name CN CN1, CN2 e CN3 North Centre NO NO1, NO2 e NO3 Northwest SE SE, SE2 e SE3 Southeast SU SU1, SU2, SU3 e SU4 South SO SO1, SO2 e SO3 Southwest T Neves/ Salim F Maluf, Tietê, and MB TS, TT e PB Pinheiros-Bandeirantes LE LE1, LE2, LE3 e LE4 East 62 of 134 C - Diagnosis GET DET Name OE OE1, OE2, OE3, e OE4 West Source: CET. Traffic Control Centrals Due to the city dimensions, traffic control and monitoring activities are also decentralized and divided into various centrals, such as Operational Control Centrals (CCOs), Traffic Light Control Departments (DCSs), among others. ▪ Traffic Light Control Department Each Traffic Engineering Management unit (GET) in São Paulo is associated with a DCS, a department responsible for centralized traffic lights and monitoring systems. They operate in the same area, but with different attributions (Annex C5. Management Areas Map and Figure 20). Information about identification, address and activities of each DCS was provided by the Terms of Reference of the Modernization and Maintenance Tender (2019): T ABLE 17. I DENTIFICATION AND LOCATION OF TRAFFIC LIGHT C ONTROL D EPARTMENTS (DCSS) DCS GET CTA (old name) DCS Address DCS-CN 7 CN CTA 2 Marques de São Vicente Avenue, 2154 DCS-NO 2 NO CTA 2 Marques de São Vicente Avenue, 2154 DCS-SE 3 SE CTA 3 Francisco Marengo Street, 1980 DCS-SU 4 SU CTA 4 Nina Rodrigues Square, s/nº DCS-SO 1 SO CTA 1 Bela Cintra Street, 385 DCS-MA 6 MB CTA 6 Nações Unidas Avenue, 7203 DCS-LE 8 LE - Américo Salvador Novelli Street, 88 DCS-OE 5 OE CTA 5 Sumidouro Street, 546 Source: CET.  DCS-CN 7: This Traffic Light Control Department is responsible for centralizing traffic control and monitoring systems that cover the centralized traffic lights and closed-circuit television (CCTV) networks for GET-CN (North Centre).  DCS-NO 2: This Traffic Light Control Department is responsible for centralizing traffic control and monitoring systems that cover the centralized traffic lights and closed-circuit television (CCTV) networks for GET-NO (Northwest).  DCS-SE 3: This Traffic Light Control Department is responsible for centralizing traffic control and monitoring systems that cover the centralized traffic lights and closed-circuit television (CCTV) networks for GET-SE (Southeast).  DCS-SU 4: This Traffic Light Control Department is responsible for centralizing traffic control and monitoring systems that cover the centralized traffic lights and closed-circuit television (CCTV) networks for GET-SU (South).  DCS-SO 1: This Traffic Light Control Department is responsible for centralizing traffic control and monitoring systems that cover the centralized traffic lights and closed-circuit television (CCTV) networks for GET-SO (Southwest).  DCS-MA 6: This Control and Monitoring Department is located together with GET-MB and is responsible for the operational management of Marginal Pinheiros, Marginal Tietê, 63 of 134 C - Diagnosis Bandeiranes Avenue, and other roads that create the Mini Ring Road. In addition, it controls CCTV monitoring system and Variable Message Signs (PMV).  DCS-LE 8: This Traffic Light Control Department is responsible for centralizing traffic control and monitoring systems that cover the centralized traffic lights and closed-circuit television (CCTV) networks for GET-LE (East).  DCS-OE 5: This Traffic Light Control Department is responsible for centralizing traffic control and monitoring systems that cover the centralized traffic lights and closed-circuit television (CCTV) networks for GET-OE (West). According to CET’s Technical Note nº 233 [30], between 1993 and 1997 five Traffic Centrals by Area (CTAs) were implemented in São Paulo by different manufacturers: Siemens (CTA 1); PEEK (CTAs 2 and 5); and Telvent (CTAs 3 and 4). In addition, Siemens and PEEK adopted SCOOT system, while Telvent opted for ITACA. This configuration resulted in dependency problems, since traffic controllers and replacement equipment had to be compatible with their respective central. As an example, all controllers installed at the time were from the same manufacturer as the central: T99R and T400 from Siemens; TSC3 from PEEK; and RMXY from Telvent. The causes that led to this situation were, according to CET:  (a) Protocols for controller/central communication were closed and proprietary.  (b) Operational features were specific to each manufacturer. As a result, CET started to search for solutions that allowed controllers from different manufacturers to be centralized with open and standardized communication protocols. Therefore, to enable interchangeability between controllers and centrals, UTMC2 and NTCIP communication protocols were adopted in 2014. Recently, the Modernization and Maintenance Tender (2019) included SCOOT software update for adaptive traffic control in São Paulo, as well as cameras and controllers compatible with this technology. Thus, even though this tender is suspended, CET has adopted SCOOT as the standard system for São Paulo and future implementations in traffic control. Regarding operation, CET’s Technical Bulletin nº 38 [31] describes that there is a hierarchy of operation modes in a CTA. When all resources are available and the system is working at its full capacity, traffic control should be in real time. This should be the regular operation mode in a CTA. One level below the operation hierarchy is the central control with fixed time, in which central computers operate traffic lights with pre-established plans. These plans can be activated by a timetable or by an operator in the control central. In both cases there is communication between central and field equipment, thus it is possible to monitor traffic conditions and make changes in the traffic light programming, as well as supervise equipment status. Considering that all controllers are currently operating in fixed time, it can be assumed that CET’s centrals are functioning according to the second option. 64 of 134 C - Diagnosis F IGURE 20. MANAGEMENT AREAS MAP Source: Future Mobility for São Paulo. 65 of 134 C - Diagnosis Another level below this hierarchy, the next operation mode is local control with fixed time. In this case, traffic light programming is determined by the controller, which has its own plans. These plans can be activated by the controller’s timetable or by an operator on field. Local control occurs when there is a failure which prevents communication between central and controller, or when – for some special reason – an operator from control central decides to activate local plans while preserving remote traffic monitoring. As a last resort, there is manual control – field operator will activate manually the change of traffic light stages on site – when real-time or fixed time operation are not sufficient during an unexpected problem. ▪ Central for Traffic Light Maintenance and Central of Operations Management For traffic light maintenance there is a separate central, operated by CET’s Signaling Engineering Management (GSI). The Central for Traffic Light Maintenance (CMS) is located at Nações Unidas Avenue, nº 7163 and is responsible for monitoring the status of all centralized traffic lights in São Paulo. For corrective maintenance, CMS works together with the Central of Operations Management (CGO), which integrates all traffic operation and maintenance actions in the city.  CMS: This central is located at Nações Unidas Avenue, nº 7163, and it monitors all centralized traffic lights in the city. It also has access to monitoring cameras from other CET departments. After confirming an occurrence, CMS is responsible for dispatching technical teams for maintenance services.  GCO: This central is located at Bela Cintra Street, nº385, together with DCS-SO 1. Its role is to track occurrences on the road network – through monitoring systems, answering to users’ calls, and/or though SPTrans’ buses notification – and forward notifications to the responsible sectors, like CMS. Furthermore, GCO is responsible for disseminating information to media vehicles and coordinating the CCOs. ▪ Operational Control Centrals In addition to traffic light control centrals, Operational Control Centrals (CCO) are responsible for the control and operation of tunnels and its respective areas of influence, according to the following characteristics:  CCO-Ayrton Senna Tunnel: This Control Central is located at Antonio Joaquim de Moura Avenue and is responsible for the operational management of Ayrton Senna Tunnel (under Ibirapuera Park), Tribunal de Justiça Tunnel (Juscelino Kubistcheck Avenue, under Santo Amaro Avenue), Max Feffer Tunnel (Cidade Jardim Avenue, under Brigadeiro Faria Lima Avenue), Jornalista Fernando Vieira de Mello Tunnel (Rebouças Avenue, under Brigadeiro Faria Lima Avenue), and Jânio Quadros/Sebastião Camargo Tunnels (under Pinheiros River). It controls CCTV monitoring systems, Variable Message Signs (PMV), vehicle detection system, ventilation system, fire detection system and others related to tunnel management.  CCO-Maria Maluf Tunnel: This Control Central is responsible for the operational management of Maria Maluf Tunnel (link between Tancredo Neves Avenue/Afonso D’Escragnolle Taunay Avenue/Bandeirantes Avenue). It controls CCTV monitoring systems, Variable Message Signs (PMV), vehicle detection system, ventilation system, fire detection system and others related to tunnel management. 66 of 134 C - Diagnosis Traffic Light System In general, São Paulo’s traffic light system is composed of 6,489 signalized intersections, in which 5,886 are regular traffic lights and 603 are flashing yellow traffic lights for warning (Annex C6a. Traffic Light System Map and Figure 22). The vast majority (90.7%) are employed as regulatory traffic lights, controlling traffic by alternating right-of-way of various flows of vehicles and/or pedestrians. Only a small portion (9.3%) of traffic lights in São Paulo function as warning signals (flashing yellow), as it is shown in Table 18 and Figure 21. T ABLE 18. Q UANTITY OF REGULAR AND FLASHING YELLOW TRAFFIC LIGHTS IN EACH GET GET Traffic Lights % Flashing Yellow % CN 852 14% 74 12% NO 707 12% 71 12% SE 916 16% 71 12% SU 888 15% 94 16% SO 927 16% 124 21% MB 64 1% 40 7% LE 804 14% 71 12% OE 728 12% 58 10% Total 5,886 100% 603 100% Source: Traffic Lights PPMI (2018). F IGURE 21. P ROPORTION OF REGULAR AND FLASHING YELLOW TRAFFIC LIGHTS Flashing Yellow 9% Regular Traffic Lights 91% Source: Traffic Lights PPMI (2018). Since this study is focused on regular traffic lights – 5,886 intersections –, the flashing yellow traffic lights will not be included in future analysis. 67 of 134 C - Diagnosis F IGURE 22. T RAFFIC L IGHT SYSTEM MAP Source: Future Mobility for São Paulo. 68 of 134 C - Diagnosis Failures In the current contract for traffic lights maintenance, a failure is described as any defect in a given traffic light infrastructure, including: controllers, UPS, signal heads, poles, cables, ducts (if existing), optical fiber (if existing), signal head for pedestrians, and so on. To every failure, CET issues a Service Order with to inform the type of maintenance service needed and location. Services Orders are classified by priority and must be concluded within a limited time period: ▪ Priority 1: Maintenance service that could offer risk to the safety of people or goods. Execution must not exceed two hours. ▪ Priority 2: Maintenance service that, if not executed, could cause Priority 1 failures. Execution must not exceed 12 hours. ▪ Priority 3: Maintenance service that, if not executed, will compromise traffic light signaling. Execution must not exceed 24 hours. Table 19 shows the correspondence between type of failures and priority classification. T ABLE 19. T YPE OF FAILURES AND PRIORITY CLASSIFICATION Description Infrastructure Priority Traffic light off Traffic lights 1 Conflicting traffic light Traffic lights 1 Traffic light with yellow focus flashing Traffic lights 1 Traffic light with phase off Traffic lights 1 Energized traffic light pole Traffic lights 1 Damaged traffic light pole Traffic lights 1 Loose protection board Traffic lights 1 UPS failure UPS 2 Controller off central mode Traffic lights 2 Controller with display failure Traffic lights 2 Controller with programming failure Traffic lights 2 Controller with facility panel failure Traffic lights 2 Controller with communication failure Traffic lights 2 Ruptured wiring Traffic lights 2 Flashing yellow off Traffic lights 2 Damaged controller wiring Traffic lights 2 Traffic light without synchronism Traffic lights 2 Damaged signal head Traffic lights 2 Broken push button Traffic lights 2 Dislocated signal head Traffic lights 2 Lack of protection board Traffic lights 2 Lack of visor Traffic lights 2 Lack of push button Traffic lights 2 69 of 134 C - Diagnosis Description Infrastructure Priority Lack of signal head Traffic lights 2 Damaged LED module Traffic lights 2 Open controller Traffic lights 2 Dislocated pole Traffic lights 2 Loose pole Traffic lights 2 Broken lock Traffic lights 2 Low wiring Traffic lights 2 Open signal head Traffic lights 2 Access box uncovered Traffic lights 2 Pedestrian crossing lighting set is damaged Pedestrian crossing 3 Communication module Traffic lights 3 Damaged pedestrian crossing lighting pole Pedestrian crossing 3 Another type of failure regarding pedestrian Pedestrian crossing 3 crossing lighting Source: CET Bidding nº 023/17. In order to calculate the distribution of maintenance teams throughout the year, the Modernization and Maintenance Tender (2019) analyzed traffic light failures between 2017 and 2018. It was highlighted that during the summer there is an increase in Priority 1 failures caused by the rainy season, so there should be a higher number of available maintenance teams from January to March. For instance, the average number of Priority 1 failures between January/18 and March/18 was 2,242, while the annual average was 1,288 failures. According to the IDOM study [28], this is due to wiring condition – the traffic light metallic cables are old and rainwater causes short circuits that disarm the electrical protection (circuit breaker). For Priority 2 and 3 failures, the Modernization and Maintenance Tender (2019) established that the number of maintenance teams would not be increased since there was no direct relation with São Paulo’s rainy season. T ABLE 20. Q UANTITY OF FAILURES BY PRIORITY (AUGUST 2017 – AUGUST 2018) Month/Year P1 P2 and P3 Total August/17 2,270 847 3,117 September/17 1,799 1,137 2,936 October/17 1,819 1,644 3,463 November/17 1,617 2,089 3,706 December/17 1,578 1,625 3,203 January/18 2,462 1,625 4,087 February/18 1,676 794 2,470 March/18 2,587 1,067 3,654 April/18 1,759 816 2,575 May/18 1,775 1,214 2,989 70 of 134 C - Diagnosis June/18 1,596 1,405 3,001 July/18 1,846 1,167 3,013 August/18 1,706 1,308 3,014 Total 24,490 16,738 41,228 Source: CET Bidding nº 01/19. F IGURE 23. Q UANTITY OF FAILURES BY PRIORITY (AUGUST 2017 – AUGUST 2018) 3000 2500 2000 1500 1000 500 0 P1 P2 e P3 Source: CET Bidding nº 01/19. As for the reasons these failures occur, CET’s Signaling/Maintenance department stated for IDOM [28] that 50% of occurrences are due to uncontrollable factors, such as vehicle collision or civil works, and these services only need 10% of maintenance resources. The other 90% of maintenance resources are distributed among three reasons: ▪ Equipment wear: 10% of occurrences are due to equipment wear, in which is necessary to replace broken parts. This type of problem causes Priority 1 failures. ▪ Power shortage: 13% of occurrences are due to power system failures, such as unbalanced voltage, distortion of the electric wave shape, voltage fluctuations, frequency fluctuations, and long- duration interruptions/over-tensions. These cause the traffic lights to go off, which is a Priority 1 failure. ▪ Theft and vandalism: 25% of occurrences are due to theft and vandalism of traffic light infrastructure, such as copper cable theft, electronic card theft with aluminum heaters, and optical fiber damage (to check if it is a metallic cable). Also results in Priority 1 failures. Therefore, in general, theft and vandalism can be considered the most aggravating reason for traffic light failures in São Paulo, as it represents 25% of all occurrences and it demands 47% of maintenance resources. 7.1.1.3 Controllers Based on the traffic light registration from 2018 (Traffic Lights PPMI), São Paulo has 4,296 controllers and 1,590 dependent intersections that are connected to a controller (Annex C7. Controllers Map and Figure 25). 71 of 134 C - Diagnosis T ABLE 21. Q UANTITY OF CONTROLLERS IN EACH GET Dependent GET Controllers Intersections CN 645 207 NO 510 197 SE 681 235 SU 642 246 SO 621 306 MB 51 13 LE 637 167 OE 510 218 Total 4,297 1,589 Source: Traffic Lights PPMI (2018). Regarding operation mode, primary data was not provided by CET. However, according to the IDOM study in 2019 [28], all intersections in São Paulo operate in fixed time because the detection network is disconnected and without maintenance. So, even if an intersection is connected to a DCS central, it will not be working in real-time mode. Still referring to the IDOM study – that had access to CET information from February 2019 – there are 232 centralized controllers in São Paulo and none operating in real time with SCOOT (Table 22). T ABLE 22. Q UANTITY OF CENTRALIZED CONTROLLERS IN EACH DCS Data/Central DCS-SO DCS-CN DCS-NO DCS-SE DCS-SU DCS-OE DCS-LE DCS-MB Total Centralized 43 27 46 0 26 82 0 8 232 Controllers Real Time Crossings 0 0 0 0 0 0 0 0 0 (SCOOT) Source: Adapted from “D1 – Status Quo Report” (IDOM, 2019). Nonetheless, São Paulo still has many functioning push buttons that communicate to the controller a demand for pedestrians to cross the street (Annex C7. Controllers Map and Figure 25). Unlike other detectors, push buttons supply and substitution are included in the maintenance contracts. Currently, 52.1% (2,240) of all controllers are actuated by push buttons. T ABLE 23. Q UANTITY OF CONTROLLERS WITH PUSH BUTTONS FOR PEDESTRIANS Push button GET Yes % No % CN 353 16% 292 14% NO 266 12% 244 12% SE 328 15% 353 17% SU 379 17% 263 13% 72 of 134 C - Diagnosis Push button GET Yes % No % SO 289 13% 332 16% MB 12 1% 39 2% LE 349 16% 288 14% OE 265 12% 245 12% Total 2,241 100% 2,056 100% Source: Traffic Lights PPMI (2018). The Traffic Lights PPMI (2018) registration also provided information about the possibility of centralization, defined by controller model and specification. As it is shown in Table 24 and Figure 24, most controllers in São Paulo cannot be centralized (68.2% of total) and the majority of non-centralized controllers are located within GET-SE, in the southeast region of the city. Regarding centralizable controllers (31.8%), 22% are located in GET-OE or west region of São Paulo (Annex C7. Controllers Map and Figure 25). T ABLE 24. Q UANTITY OF CONTROLLERS THAT ARE CENTRALIZABLE AND NON -CENTRALIZED IN EACH GET GET Centralizable % Non-centralized % CN 198 15% 447 15% NO 147 11% 363 12% SE 137 10% 544 19% SU 135 10% 507 17% SO 270 20% 351 12% MB 33 2% 18 1% LE 145 11% 492 17% OE 300 22% 210 7% Total 1,365 100% 2,932 100% Source: Traffic Lights PPMI (2018). F IGURE 24. P ROPORTION OF CONTROLLERS THAT ARE CENTRALIZABLE AND NON -CENTRALIZED Centralizable 32% Non centralized 68% Source: Traffic Lights PPMI (2018). 73 of 134 C - Diagnosis F IGURE 25. C ONTROLLERS MAP Source: Future Mobility for São Paulo. 74 of 134 C - Diagnosis Based on information from 2019, the IDOM study [28] identified 1,396 centralizable controllers – 31 more than what was indicated in 2018 – and listed the models that were not compatible with UTMC SCOOT. According to their findings, only 63.8% of all centralizable controllers can be connected with UTMC SCOOT implemented by Siemens, a total of 891 controllers. Supplier The supplier of each controller implemented in São Paulo was defined according to its model, which was given by Traffic Lights PPMI registration table (Annex C8. Controller Supplier Map and Figure 26). For models ‘ESC-2’, ‘ESC-3’, ‘ESCRAVO’, ‘S-4’, ‘SEC-2’, and ‘S-5/G’ no supplier was identified. However, these controllers only represent 0.8% of the total. To match controller supplier and model, the following correspondence was adopted: T ABLE 25. Q UANTITY OF CONTROLLERS BY SUPPLIER IN EACH GET Controller Model Quantity Supplier DATAPROM DP-40 276 Dataprom DATAPROM DP-40 FX 143 CD-100 9 CD-200 599 Digicon CD-300 66 DIGICON FCA 503 GREENWAVE-FX 798 Greenwave GREENWAVE-TR 533 FLEX III 350 FLEX III/A 47 Newtesc FLEX IV 106 PTC1 148 PTC1-5 2 PEEK Traffic TSC3 84 SERTTEL FX 69 Serttel SERTTEL TR 4 T-400 101 T-800 23 Siemens T99-1 74 ST-900 74 RBY 3 RBY-4 3 RMX-Y3 94 Telvent RMX-Y4 104 RMY-3 17 75 of 134 C - Diagnosis Controller Model Quantity Supplier RMY-4 29 ESC-2 2 ESC-3 1 ESCRAVO 15 S-4 2 No data SEC-2 11 S-5/G 1 No model 5 Source: Traffic Lights PPMI (2018). In general, Greenwave has the largest presence in São Paulo, representing 31% of all implemented controllers. Created in 2013, Greenwave is headquartered in the city of Osasco (located in the SPMR) and provides both fixed time and real-time controllers for CET – ‘GREENWAVE-FX’ and ‘GREENWAVE-TR’ models respectively. After Greenwave, approximately 27.4% of all controllers belong to Digicon, another national company from the city of Gravataí (State of Rio Grande do Sul). Together with Newtesc and Dataprom, these four Brazilian companies represent almost 80% of the current infrastructure. T ABLE 26. Q UANTITY OF CONTROLLERS BY SUPPLIER IN EACH GET PEEK No GET Dataprom Digicon Greenwave Newtesc Serttel Siemens Telvent Traffic data CN 94 172 194 56 33 23 69 0 4 NO 29 217 167 47 28 12 7 0 3 SE 68 199 230 100 0 0 0 73 11 SU 85 197 212 56 0 14 17 58 3 SO 23 111 171 160 79 13 37 21 6 MB 8 5 11 4 1 0 0 22 0 LE 89 189 201 44 0 11 20 76 7 OE 23 87 145 36 93 0 122 0 4 Total 419 1,177 1,331 503 234 73 272 250 38 Source: Traffic Lights PPMI (2018). 76 of 134 C - Diagnosis F IGURE 26. C ONTROLLER SUPPLIER MAP Source: Future Mobility for São Paulo. 77 of 134 C - Diagnosis F IGURE 27. P ROPORTION OF CONTROLLERS BY SUPPLIER Telvent No data Siemens 6% 1% Serttel 6% 2% Dataprom 10% PEEK Traffic 5% Digicon Newtesc 27% 12% Greenwave 31% Source: Traffic Lights PPMI (2018). Age Traffic controller’s date of implementation was not provided by CET, therefore all information regarding equipment age will be based on the IDOM study [28]. F IGURE 28. P ROPORTION OF CONTROLLERS BY AGE More than 15 years 1% 10 to 15 years 24% 0 to 5 years 52% 5 to 10 years 23% Source: Adapted from “D1 – Status Quo Report” (IDOM, 2019). In February 2019, the average age of controllers in São Paulo was 8.5 years. In general, 75% of all controllers had less than 10 years, while the majority (52% of total amount) had between 0 and 5 years. The oldest models operating at that time were ‘CD-100’ and ‘T99-1’ with 25 years old. Both models represent approximately 1% of all implemented controllers and are no longer manufactured. Protocols When CET implemented traffic light control centrals in São Paulo in the 90’s, there was an exclusivity dependency issue since all controllers should be compatible with the central’s technology. This led CET to buy controllers and equipment from the same manufacturers that installed the CTAs (Siemens, PEEK, and 78 of 134 C - Diagnosis Telvent). To allow other manufacturers to provide controllers for São Paulo, CET started studying the adoption of open and standardized communication protocols to enable interoperability. As a result, SMT published Ordinance nº 02/14 in January 2014 to establish communication protocols for Intelligent Transportation Systems (ITS), including traffic monitoring, surveillance, controllers, and control centrals implemented in São Paulo. For real-time traffic controllers, it determines that communication between central and controllers should adopt protocols UTMC (Urban Traffic Management Control) Type 2 or NTCIP (National Transportation Communication for ITS Protocol). The same applies for fixed time/actuated traffic controllers, which should also install an XML complement to enable: ▪ Transmission of one or more signal plans to the controller. ▪ Transmission of timetable to the controller. ▪ Activate or release signal plans/modes. ▪ Confirmation that the controller received parameters sent by the central. ▪ Status monitoring. ▪ Failure monitoring. 7.1.1.4 Detectors Traffic detection is essential for real-time operation of traffic lights, however there are no working detectors in São Paulo – according to IDOM study [28]. Since all sensors are disconnected, intersections operate in fixed time. This situation is aggravated by a lack of maintenance and equipment substitution, which are not contemplated in the current maintenance contract. According to CET’s technical specifications, there are two types of sensors that can be implemented in São Paulo: ▪ Overhead detectors: it is any vehicular detection system that does not depend on floor sensors implementation (e.g., virtual detectors). This detection system must be able to provide flow and occupation data. ▪ Magnetic detector: it is any vehicular detection system that, installed under the pavement, has autonomy in its functioning. Therefore, the magnetic sensor must not have any type of physical connection to power sources, controllers, or any other external equipment. As virtual detectors and inductive loops are the most relevant examples of vehicular sensors to CET, both have individual technical specifications: ▪ Virtual detector (camera): must consist of a video detection camera and an interface that allows monitoring of up to 4 lanes. The cameras must allow vehicle counting, measurement of occupied zone, advance detection, among others to replace induction loops. ▪ Inductive detector (loop): must consist of electronic boards and inductive loops installed in a specific section of the road, capable of detecting vehicular flow. Unfortunately, information about quantity, type, or location of traffic detectors in São Paulo is unavailable. As an alternative, the Modernization and Maintenance Tender (2019) can provide some insights about what CET expects from a future detection system in the city (Table 27). 79 of 134 C - Diagnosis T ABLE 27. Q UANTITY OF DETECTORS SUPPLY IN THE M ODERNIZATION AND MAINTENANCE TENDER (2019) Detector Lot 1 Lot 2 Lot 3 Lot 4 Inductive loop 239 units 333 units 194 units 184 units Virtual loop (camera) 1,079 units 1,500 units 879 units 829 units Source: CET Bidding nº 01/19. 7.1.2 Public Transport Diagnosis São Paulo’s current public transport system is held by two power spheres: the city hall (city bus); and the state government (subway, train, and intercity bus). In addition, private companies participate in this arrangement providing services to the public authorities. Despite the heterogeneity of active institutions, São Paulo’s public transport network works in an integrated manner thanks to Bilhete Único – a unified payment system for municipal bus, subway and urban train fares. With Bilhete Único, the value of one fare allows up to 4 bus trips in 3 hours, in addition to offering a discount when transferring modals to subway/train. As previously indicated, a large portion of trips originated in São Paulo are made by collective modes of transport, such as bus, subway, and train. Given its importance for the mobility dynamics of São Paulo residents, the main means of public transport will be discussed below. 7.1.2.1 Subway The first subway line in São Paulo was inaugurated in 1974, connecting Jabaquara and Vila Mariana neighborhoods – a section of the future 1-Blue line. Almost 50 years later, São Paulo’s subway network – also known as metrô – has more than 101 kilometers in length, 6 operating lines and 89 stations (Annex C9. Subway and Train Network Map and Figure 30). From these, 4 lines are operated by Metrô (São Paulo Subway Company), while lines 4-Yellow and 5-Lilac are operated by private companies under a public- private concession. Regarding the metrô lines, we can highlight: ▪ Line 1-Blue: It was the first subway line to be built in Brazil, also called the North-South Line due to its orientation in the city. It is 20.2 km long and has 23 stations. It currently has the second largest number of passengers, only behind Line 3-Red. ▪ Line 2-Green: It was inaugurated in 1991 under the name of “Linha Paulista”, with a sec tion built along São Paulo’s financial center located on the homonymous avenue. It is 14.7 km long and has 14 operational stations. In 2021, an extension project for the line was announced – an additional 8.4 km and 8 new stations, connecting it with Line 3-Red. ▪ Line 3-Red: Inaugurated in 1979, Line 3-Red – formerly known as East-West Line – was the second to be built in São Paulo. With 22 km in length and 18 stations, it serves neighborhoods located in the extreme east zone of the city and transports the highest number of passengers in the entire system. ▪ Line 15-Silver: Also known as Monorail, the first section of the Line 15-Silver was inaugurated in 2014. It currently has 11.6 km in length and 10 operational stations, but when completed it will have 26.6 km and 18 stations. ▪ Line 4-Yellow: Operated by Via Quatro company under a 30-year concession contract, Line 4- Yellow was inaugurated in 2010 and is still under construction. Today, it is 11.4 km long and has 10 operational stations. It was designed to serve the Southwest urban growth vector and the new financial center of Faria Lima, crossing the richest areas of the city. 80 of 134 C - Diagnosis ▪ Line 5-Lilac: The first section was inaugurated in 2002, connected only to CPTM's Line 9-Emerald. A peculiarity of this section was its insertion in territories of precarious urbanization and high levels of violence, something unusual in Metrô projects. It was only in 2018 that Line 5-Lilac was integrated into the subway network, with the construction of a second section connecting it to Line 1-Blue and 2-Green. It currently has 19.9 km of extension, 17 stations and is operated by the concessionaire company Via Mobilidade. Despite being quite extensive in length, São Paulo’s Metrô network covers only a portion of the urban territory, with a high concentration of lines and stations in the richest regions of the city. In addition, civil works on new sections and lines are constantly delayed or suspended for various reasons. F IGURE 29. D IAGRAM OF SÃO PAULO ’S METROPOLITAN TRANSPORT Source: Metrô. 81 of 134 C - Diagnosis F IGURE 30. S UBWAY AND TRAIN NETWORK MAP Source: Future Mobility for São Paulo. 82 of 134 C - Diagnosis As in other cities around the world, the Covid-19 Pandemic had a great impact on urban mobility in São Paulo. After the World Health Organization (WHO) declared it a global pandemic in March/2020, São Paulo State Government decreed the temporary closure of all services considered non-essential, in addition to the suspension of school classes and events with more than 500 people. As a result, there was a sharp drop in transported passengers by Metrô between the months of February and April 2020. Among the various waves of contamination and measures to restrict/open stores, the flow of passengers returned to grow from the second half of 2020 onwards, but reaching only half the demand when compared to numbers before the pandemic. As it is shown in Figure 31, passenger flow in February 2021 represents 52% of what it was in February 2020 (one month before WHO’s declaration of global Covid-19 Pandemic). F IGURE 31. F LOW OF PASSENGERS TRANSPORTED BY SUBWAY (J ANUARY 2020 – MARCH 2021) 140,000,000 120,000,000 Transported Passengers 100,000,000 80,000,000 60,000,000 40,000,000 20,000,000 0 Jul-20 Apr-20 May-20 Jun-20 Aug-20 Sep-20 Nov-20 Jan-20 Feb-20 Mar-20 Dec-20 Jan-21 Feb-21 Mar-21 Oct-20 Source: Metrô, Via Quatro and Via Mobilidade. Lastly, São Paulo Metrô operates together with the urban rail network through free integration stations, where passengers can transfer lines without an additional fare. With this integration, mass transport starts to operate at the metropolitan level, connecting neighboring cities and more distant neighborhoods to the São Paulo subway network. T ABLE 28. P ASSENGER ENTRY BY LINE IN MARCH 2021 Company Line Passengers (thousands) Line 1 – Blue 9,795 Line 2 – Green 5,570 Metrô Line 3 – Red 12,699 Line 15 – Silver 887 Via Quatro Line 4 – Yellow 2,323 Via Mobilidade Line 5 – Lilac 3,447 Line 7 – Ruby 4,750 Line 8 – Diamond 4,923 CPTM Line 9 – Emerald 4,938 Line 10 – Turquoise 4,248 83 of 134 C - Diagnosis Company Line Passengers (thousands) Line 11 – Coral 7,210 Line 12 – Sapphire 3,531 Line 13 – Jade 202 Source: CPTM, Metrô, Via Quatro and Via Mobilidade. 7.1.2.2 Metropolitan Train Operated by CPTM (São Paulo Metropolitan Train Company), São Paulo’s urban train network cover 23 municipalities in Jundiaí Urban Agglomeration and the Metropolitan Region of São Paulo. It is an extremely important modal for maintaining long-distance commuting in the metropolitan region, since all train lines pass through São Paulo. Today, CPTM has a total of 94 operational stations and 271 km of tracks, divided into 7 lines. Except for Line 13-Jade, all lines have a transfer station from CPTM to São Paulo Metrô (Annex C9. Subway and Train Network Map and Figure 30). About the lines, we can highlight: ▪ Line 7-Ruby: Its origins date from 1867 as the first railway in the State of São Paulo, named “São Paulo Railway Company Ltd.”, which connected Jundiaí to Santos’ Port. It currently connects cities in the Jundiaí Urban Agglomeration (specifically, Jundiaí, Várzea Paulista and Campo Limpo Paulista municipalities) to São Paulo city center. It is CPTM's largest line, with 19 stations in operation and 62.7 km in length. ▪ Line 8-Diamond: It was inaugurated in 1875 as part of the Sorocabana Railroad and currently connects Barueri, Carapicuíba, Itapevi, Jandira and Osasco municipalities to São Paulo city center. It is 41.6 km long and has 24 operational stations. In April 2021, CCR business group won a 30-year concession contract for the line. ▪ Line 9-Emerald: Inaugurated in 1957 as part of the Sorocabana Railroad, the Emerald Line connects Osasco municipality to the Grajaú district in São Paulo. It has 18 stations in operation and is 31.8 km long, with expansion works in progress for Varginha municipality. In April 2021, CCR business group won a 30-year concession for the line. ▪ Line 10-Turquoise: Just like Line 7-Ruby, the Turquoise Line comprises a section of the former “São Paulo Railway Company Ltd.”, inaugurated in 1867. Currently it connects Mauá, Ribeirão Pires, Rio Grande da Serra, Santo André and São Caetano do Sul municipalities to São Paulo city center. It has 13 operational stations and is 37.2 km long. ▪ Line 11-Coral: Inaugurated in 1875 as the Northern Railroad,Line 11-Coral was also known as “Expresso Leste” or East Express. It currently connects Ferraz de Vasconcelos, Mogi das Cruzes, Poá and Suzano municipalities to São Paulo city center. In addition, it serves several neighbourhoods located in the far east zone of São Paulo that the subway network does not reach. It is 50.7 km long and has 16 stations in operation. ▪ Line 12-Sapphire: Built as a variant of “Central do Brasil” Railroad in 1934, the Sapphire Line connects Poá and Itaquaquecetuba municipalities to São Paulo city center. It has 13 stations and is 38.9 km long. ▪ Line 13-Jade: It is the most recent line and the first to be built entirely by CPTM. Inaugurated in 2018, Line 13-Jade connects São Paulo city center to São Paulo’s International Airport, located in Guarulhos municipality. It is 10.3 km long and has 3 operational stations. In the same way that São Paulo’s subway network was affected by Covid-19 Pandemic, the metropolitan train network presented a large drop in users due to social distancing measures. In general numbers, the 84 of 134 C - Diagnosis integrated report “Relatório Integrado 2020” published by Metrô registered 504.6 mil lion passengers transported by CPTM in 2020, which indicates a decrease of 42% compared to 2019. 7.1.2.3 City Bus The operation of city buses in São Paulo is done through private companies that report directly to SPTrans (São Paulo Transport). As the main mean of collective transport, buses correspond to 21% of the total trips originated in São Paulo, with a higher share than other means of public transport (trips by subway – 11%; trips by train – 3%). The intensity in bus use in São Paulo is due to its capillarity in neighborhoods and local roads, covering regions not served by mass transport (Annex C10b. City Bus Network Map – Bus Lines and Figure 33). In addition, unlike the subway and metropolitan train, buses in São Paulo operate during the night (between 0:00AM and 04:00AM) with the “Noturno” operation, being the only means of public t ransport functioning 24 hours a day. Also, São Paulo has exclusive spaces for bus traffic, an initiative that began to be implemented in the 70s. Currently, there are 500km of exclusive lanes and 131.2km of bus corridors that allow a more efficient flow of buses in the city (Annex C10a. City Bus Network Map – Bus corridors and Bus Exclusive Lanes and Figure 32). Furthermore, several bus corridors and exclusive lanes are connected to subway and train stations, demonstrating the complementarity of these modals and their integrated dynamics. Regarding the city bus system, we can highlight: ▪ Registered fleet of 14,021 buses, of which 55% have air conditioning and 8% have free WiFi. ▪ Operation of 1,336 city bus lines (Annex C10b. City Bus Network Map – Bus Lines and Figure 33). ▪ A total of 116 million passengers transported in the month of March/2021. ▪ 32 city terminals, which are arrival or departure points for city buses. City terminals have a structure to assist passengers, such as charging stations for the Bilhete Único card. ▪ More than 21 thousand bus stops throughout the city. Each bus stop has a poster with visual information about the lines and operating hours. For monitoring and management, SPTrans has an independent Operations Central (COP) and two main operational systems: ticketing and the Integrated Monitoring System (SIM). Ticketing information is currently used for Bilhete Único fare division between companies, but also for planning and operation. Due to internal security procedures, ticketing data is stored in each bus and later uploaded locally at the garage. This process delays data availability in the system up to 8 days14. On the other hand, SIM is a management software that monitors the quality of service provided by operating companies through field data collecting. Currently, all city buses are equipped with an Automatic Vehicle Location (AVL) that informs its position every 45 seconds, providing real-time data. This data is compared to Infotrans database – provides information on bus lines, timetable, itinerary, and so on – to evaluate if operation is adequate. 14 Information provided during an internal meeting with SPTrans in August 2021. 85 of 134 C - Diagnosis F IGURE 32. C ITY BUS NETWORK MAP – BUS CORRIDORS AND BUS E XCLUSIVE L ANES Source: Future Mobility for São Paulo. 86 of 134 C - Diagnosis F IGURE 33. C ITY BUS NETWORK MAP – BUS L INES Source: Future Mobility for São Paulo. 87 of 134 C - Diagnosis In addition to managing concession companies and bus lines, SIM real-time data on bus location is offered as an online consultation service named “Olho Vivo”, which is aimed at citizens that want to plan their bus trip in São Paulo. Created in 2005 for bus fleet inspecti on, “Olho Vivo” system was expanded in 2012 to provide the user with information on itinerary, speed, and travel time in the main corridors of the city [32]. Currently, it is possible to access this monitoring system through the “Olho Vivo” web page and search for information in three modalities: “De olho na linha”, which allows for the line consultation through an identification code and shows real-time position of the buses that serve it; “De olho no ponto”, which allows you to identify in real time the buses that are approaching the chosen bus stop and all lines that serve it; and “De Olho na Via”, which shows the performance of São Paulo’s main corridors and the average speed of buses circulation. SPTrans also provides bus fleet real-time information through the “Olho Vivo” API, open data used by researchers and developers of mobility apps. Even though there is no official application for public transport, SPTrans is interested in developing an integrated app for mobility including different types of means of transport and payment methods – initiative closely related to MaaS ideas. F IGURE 34. F LOW OF PASSENGERS TRANSPORTED BY THE CITY BUS FLEET (J ANUARY 2020 – MARCH 2021) 250,000,000 Transported Passengers 200,000,000 150,000,000 100,000,000 50,000,000 0 Jul-20 Jun-20 Jan-20 Apr-20 May-20 Aug-20 Nov-20 Dec-20 Mar-21 Feb-20 Mar-20 Sep-20 Jan-21 Feb-21 Oct-20 Source: SPTrans. In the same way that Covid-19 Pandemic affected São Paulo's subway network, the total number of passengers transported by city buses dropped 65% between February and April 2020. Similar to the subway's recovery curve, bus passengers’ volume grew in the second half of 2020, stabilizing at approximately 120 million passengers per month. However, even with a decrease in the number of passengers, complaints about the occupancy and waiting time for city buses remained high due to the active fleet reduction during the pandemic. According to “Viver em São Paulo” opinion poll carried out in September 2020, 53% of respondents believes that the waiting time for buses has increased compared to what it was a year ago. Simultaneously, 49% of respondents feel that bus occupancy has increased in the same time period [19]. The survey also asked specific questions for respondents who choose other means of transport rather than the city bus. When questioned about why not using the bus as a means of transport in São Paulo, respondents pointed out the fear of contracting the coronavirus and bus occupancy as the main reasons - both with 35% of mentions in the survey [19]. 88 of 134 C - Diagnosis Regarding the relation between São Paulo’s bus network and traffic lights, georeferencing tools were used to locate signalized intersections that are implemented in bus corridors and exclusive lanes (Annex C10a. City Bus Network Map – Bus corridors and Bus Exclusive Lanes and Figure 32). Currently, only a small portion of traffic lights are located in bus corridors (11.1% of total). However, based on expansion plans from São Paulo’s Master Plan (2014) and Mobility Plan (2015), this number can increase to 35.5% until 2025 due to the implementation of new bus corridors. As for exclusive lanes, 33% of traffic lights are implemented in roads with exclusive lanes for buses. T ABLE 29. Q UANTITY OF SIGNALIZED INTERSECTIONS IN BUS CORRIDORS (EXISTING AND PLANNED ) Bus corridor GET Existing % Planned % No corridor % CN 31 5% 226 16% 595 16% NO 132 20% 98 7% 477 13% SE 44 7% 130 9% 742 20% SU 104 16% 260 18% 524 14% SO 220 34% 154 11% 553 15% MB 6 1% 52 4% 6 0% LE 14 2% 363 25% 427 11% OE 100 15% 158 11% 470 12% Total 651 100% 1,441 100% 3,794 100% Source: Traffic Lights PPMI (2018) and SPTrans (2017). F IGURE 35. P ROPORTION OF SIGNALIZED INTERSECTIONS IN BUS CORRIDORS (EXISTING AND PLANNED ) Existing 11% Planned 25% No corridor 64% Source: Traffic Lights PPMI (2018) and SPTrans (2017). 89 of 134 C - Diagnosis T ABLE 30. Q UANTITY OF SIGNALIZED INTERSECTIONS IN EXCLUSIVE LANES FOR BUSES Exclusive lane GET Yes % No % CN 257 14% 595 15% NO 172 9% 535 13% SE 350 19% 566 14% SU 273 15% 615 15% SO 134 7% 793 19% MB 30 2% 34 1% LE 353 19% 451 11% OE 247 14% 481 12% Total 1,816 100% 4,070 100% Source: Traffic Lights PPMI (2018) and SPTrans (2016). F IGURE 36. P ROPORTION OF SIGNALIZED INTERSECTIONS IN EXCLUSIVE LANES FOR BUSES Yes 31% No 69% Source: Traffic Lights PPMI (2018) and SPTrans (2016). When crossing data about signalized intersections in public transport corridors (bus corridors and exclusive lanes) and possibility of centralization, it is possible to observe in Table 31 that most traffic lights located at bus corridors are centralizable (370 intersections out of 651 or 56.8%). On the other hand, more than half traffic lights in exclusive lanes cannot be centralized (858 intersections out of 1,624 or 52.8%) – albeit it is by a small margin. T ABLE 31. Q UANTITY OF SIGNALIZED INTERSECTIONS CONSIDERING POSSIBILITY OF CENTRALIZATION AND PUBLIC TRANSPORT CORRIDORS . Bus corridor Exclusive lane No corridor Total Centralizable 370 766 988 2,124 90 of 134 C - Diagnosis Bus corridor Exclusive lane No corridor Total Non-centralized 281 858 2,623 3,762 Total 651 1,624 3611 5,886 Source: Traffic Lights PPMI (2018) and SPTrans (2016). 7.1.2.4 Individual public transport Taxi Alternatives Uber arrived in Brazil in May 2014, starting its operation in the city of Rio de Janeiro. In June, Uber drivers were already circulating in São Paulo with a reduced fleet for a test phase. As in other cities around the world, Uber activities in São Paulo provoked angry reactions from taxi drivers who argued that this new mean of transport was illegal. Their protests gained strength after Decree nº 56.981/2016 publication, which regulated individual transport via application in São Paulo. Despite taxi driver s’ protests, Uber continued to grow and, according to data from 2019, the city has around 150,000 registered drivers. Uber has different travel models to be chosen. In São Paulo, the available models are: exclusive trips with luxury cars (Uber Black); a more comfortable experience (Uber Comfort); trips with popular cars (UberX); cars with luggage space (Uber Bag); trips with the possibility of including pets (Uber Pet); shared rides (Uber Juntos); and – for a brief period in 2016 and again in 2019 – air taxi services (UberCopter). Uber also expanded its services to deliver food from restaurants (Uber Eats). Due to Covid-19 pandemic, Uber sought to ensure the safety of users and drivers with: suspension of Uber Juntos (travel sharing mode) to avoid crowding; contactless delivery option, with new "Leave at the door" feature in which the user avoids direct contact with the delivery person; mandatory use of masks for drivers and users; and measures such as open windows and vehicle sanitation. Also, Uber Flash modality was launched so that users could request the delivery of personal items without the need to travel. Currently, Uber fare is defined by five criteria, which are: base fare (starting price of the trip, fixed amount); minimum price; cost per minute; cost per kilometer; and booking fee. Taking as an example the values for UberX category, the most popular, base price is BRL 1.80; minimum price is BRL 6.29; cost per minute is BRL 0.18; and cost per kilometer is BRL 1.34. Values may vary according to city region, demand and time. With a similar approach, 99 is a Brazilian startup founded in 2012 as “99 Taxi”, which emerged as an application to connect taxi drivers and passengers. To compete with companies like Uber, 99 launched 99POP service for private drivers in 2016. In early 2018, the company was acquired by Chinese Didi Chuxing and became the first Brazilian unicorn (startups valued at US$1 billion or more). Currently, 99 works with four categories. In addition to 99POP and 99Taxi already mentioned, there is also 99Comfort category for more comfortable trips, and 99Top, an exclusive taxi service. 99 also works with food delivery services through 99Food and delivery of personal items with 99Entregas. Much like Uber, 99 rate is charged with a starting price, minimum fare, cost per kilometer and cost per minute. To compete with Uber and 99, Cabify – a Spanish startup – arrived in Brazil in 2016 and started its operation in the city of São Paulo. In 2017, Cabify acquired Easy Taxi, a Brazilian app for taxi hailing. However, after five years of operation, the company ended its activities in Brazil due to the current health situation and socioeconomic crisis caused by Covid-19 pandemic. Cabify highlighted that it will continue 91 of 134 C - Diagnosis its operation in Latin America and Spain, where cities had a good recovery rate in demand – around 75% – which did not happen in São Paulo. Car Sharing Car sharing is a comprehensive term, including various types of operation. According to Deloitte’s report “Car Sharing in Europe”, there are three distinct business models: free-floating car sharing, it allows customers to pick up and return the vehicle anywhere within a certain area; stationary car sharing, it has fixed stations and (usually) provides only round trips with the start and end points being the same; and peer-to-peer car sharing, it offers vehicles belonging to private individuals to a specific user community. In São Paulo, stationary, and peer-to-peer models were identified as alternatives to the traditional “re nt a car” services. Turbi is a stationary car sharing startup created in 2017 that currently has more than 400 stations in São Paulo Metropolitan Region, including parking lots, airports, office, and residential buildings. In July 2020, Turbi had 700 vehicles in operation, with plans to expand its fleet to 2,000 cars by the end of 2020. However, the company decided to postpone investments due to Covid-19 pandemic. Moobie is a peer-to-peer car sharing startup, also created in 2017, that connects users who need a car and those who want to rent their private vehicle. In 2020, meeting the demand of users, Moobie launched a monthly rental modality for BRL 999.00 per month in which multiple car options are available. Nonetheless, the company suspended operations in May 2021 due to the pandemic advance, with no scheduled return date. Kinto Share, Toyota’s mobility division (Toyota Mobility Services) present in Brazil since 2020, offers a car rental service through Kinto Share Brazil app. This service is offered in six Brazilian state capitals, including São Paulo and its metropolitan region. In their catalog, Toyota and Lexus models are available, with daily rates ranging from BRL 149.00 to BRL 750.00. Users can also choose the hourly rental modality. For both options, cars are picked up and returned at Toyota dealerships or delivered to your home as an extra service. Beepbeep is similar to other stationary car sharing services, but it operates exclusively with electric cars. In October 2020, it had 60 stations distributed throughout the cities of São Paulo, Campinas, Guarulhos, and São José dos Campos, with a fleet of 35 electric cars. Created in 2019, Beepbeep services are accessed by mobile app with a starting price of BRL 9.90, plus an amount per minute that ranges from BRL 0.15 and BRL 0.79 – for longer trips, the rate per minute is lower. Stations are located in supermarkets, parking lots, residential buildings, malls, and hotels. In addition, cars can be returned at any station. Finally, Joycar is a Brazilian startup created in 2015 that specializes in management of shared fleets for companies. Their operation consists of equipping the fleet with a hardware that controls user registration, car activity (opening doors, starting the engine, and so on), and car tracking – enabling and automating the corporate car sharing system. Joycar currently serves 20 companies that pay a monthly fee per registered car, such as Honda, Hyundai, Petrobras and Raízen. 7.1.3 Communication Diagnosis Currently, there are two municipal communication networks in São Paulo, and both are related to transport networks. Firstly, SPTrans has an optical fiber infrastructure to connect variable message signs in “Smart Corridors” and bus terminals. These variable message signs display information for bus users – approaching lines and waiting time, for example – and the fiber network can be underground or aerial 92 of 134 C - Diagnosis (overhead). To provide real-time data, city buses are equipped with an AVL module (Automatic Vehicle Location) that communicates with an Integrated Monitoring System (SIM) by GPRS every 45 seconds. A bus user can also check real-time information in SPTrans online platform “Olho Vivo”. A second existing infrastructure is CET’s communication network, destinated to data and image transmission between DCSs’ central computers and field equipment (traffic controllers, cameras, variable message signs, and so on). According to the IDOM study [28], CET’s infrastructure is composed of: ▪ Optical fiber connection between traffic centrals (traffic control, monitoring and maintenance) – Ethernet network implemented with no redundancies, susceptible to simple failures. ▪ Optical fiber connection to field equipment (controllers and cameras) – “GPON” type network, based on the FTTH network standards. It connects 190 controllers to DCSs/CTAs and 500 CCTV cameras. ▪ Metallic cable communication for traffic controllers – Copper telephone cables used for controller centralization exclusively in CTA 1/DCS SO-1. In addition, each CTA/DCS has a LAN network that is divided into individual VLAN’s for specific services. For example, there is a Traffic Light Control VLAN and a Video Surveillance VLAN implemented in every central. CET also hires private companies by public bidding to provide complementary communication services, such as GPRS and TETRA systems. The GPRS network is used for controller monitoring – failure report transmission – and communication with variable message signs in the road network. On the other hand, TETRA services are dedicated to emergency communication of CET’s signaling department. Another public communication network – that does not belong to São Paulo City Hall – is Infovia, a corporate data network implemented by Metrô. In May 2020, approximately 85 km of optical fiber had been installed in São Paulo’s subway infrastructure for management integration, IT services, equipment monitoring, and so on. Besides network modernization, Infovia’s project contemplated commercial exploitation as an alternative source of revenue. Currently, Infovia follows the specifications shown in Table 32. T ABLE 32. I NFOVIA CHARACTERISTICS Type of optical fiber Length Location Subway lines 1-Blue, 2-Green, and 3- 72 fiber 78,084.50 meters Red; Itaquera Rail Yard; Tamanduateí Rail Yard; among others. Connections between technical rooms and other smaller locations 12 fiber 2,676.00 meters (maintenance room, small rail yards, and so on). Connections between technical rooms 4 fiber 4,745,00 meters and internal areas. Source: PMI nº 10016093 (Metrô, 2021). The infovia network is based on IP/MPLS (IP Multiprotocol Label Switching) technology and was designed with 1 Main Ring and 6 Secondary Rings for distribution. The Main Ring has 40GE speed in redundant fibers and high-performance equipment connecting the Jabaquara Rail Yard, Metrô’s Data Center, Ana Rosa Station (line 1-Blue and 2-Green), Penha Station (line 3-Red), and Tamanduateí Rail Yard. 93 of 134 C - Diagnosis On the other hand, the 6 Secondary Rings have 10GE speed in fibers and high-performance equipment connecting the Main Ring to some administrative buildings, stations, rail yards and so on. In addition, there are access switches with 1GE speed in all subway stations, maintenance rooms, operational rooms and other areas located along lines 1-Blue, 2-Green, and 3-Red. In March 2021, Metrô published a PMI (Procedure for Expression of Interest) to gather studies for the privatization, operation, and maintenance of telecommunication systems utilizing São Paulo’s subway areas and infrastructure. Its aim was to receive technical analysis for existing products – such as Infovia – and possible business opportunities. Focusing on traffic light control, the following analysis will be based on CET’s duct infrastructure – information provided by Smart City PMI (2020). According to CET, there are 500 km of ducts for data and image transmission, however most have never been updated or improved. CET’s assumption is that 50% of the communication network needs some type of civil work repair, with optical fiber substitution due to damage or advanced age. As a frame of reference, the Modernization and Maintenance tender (2019) estimated the supply of over 660 kilometers of optical fiber for São Paulo’s traffic light system renovation. Since there was no available information in the framework of this study regarding connection status for each controller, the traffic light geodatabase was associated with CET’s duct infrastructure sh apefile using georeferencing tools. As a result, traffic controllers were categorized according to the infrastructure (underground or aerial) and connection type (optical fiber, copper, or both) identified by location – proximity with an existing duct. Nonetheless, it is important to stress that CET’s optical fiber registration is outdated and may not represent what is currently implemented. In fact, IDOM had the opportunity to interview CET about communication networks and it was mentioned that most copper cables are unused and only 200 km of optical fiber are in service [28]. 7.1.3.1 CET’s Communication Infrastructure CET’s current communication infrastructure (Annex C11. Duct Network Map and Figure 37) is composed of 257.4 km of aerial cabling (51.4% of total) and 242.9 km of underground ducts (48.6% of total). Most of overhead cables are only copper (112.2 km) or have both metallic pairs and optical fiber installed (106.4 km). As for underground ducts, the majority has only optical fiber (72.5 km) or both copper and fiber (69.3 km). However, according to Table 33 there are 56.1 km of ducts that could be considered “empty”, having neither metallic nor optical fiber connection implemented (23.1% of underground ducts). T ABLE 33. C OMMUNICATION INFRASTRUCTURE LENGTH BY CONNECTION TYPE Type of connection Aerial cabling % Underground duct % Copper 112,158 m 43.6% 45,136 m 18.6% Optical fiber 38,311 m 14.9% 72,465 m 29.8% Mixed (both metallic 106,361 m 41.3% 69,263 m 28.5% cable and optical fiber) No connection 589 m 0.2% 56,071 m 23.1% Total 257,418 m 100% 242,935 m 100% Source: Smart City PMI (2020). 94 of 134 C - Diagnosis F IGURE 37. D UCT NETWORK MAP Source: Future Mobility for São Paulo. 95 of 134 C - Diagnosis To understand how each controller is connected to a DCS central, the traffic light geodatabase was associated with CET’s communication network. According to location, it was identified that 1,417 controllers are covered by the existing infrastructure (33% of all controllers), 798 controllers covered by underground ducts and 619 controllers covered by aerial cabling. A total of 2,879 controllers (67% of total) are not covered by any type of communication infrastructure, whether they are underground or overhead. T ABLE 34. Q UANTITY OF CONTROLLERS BY COMMUNICATION INFRASTRUCTURE GET Underground ducts % Aerial cabling % No infrastructure % CN 142 18% 87 14% 416 14% NO 78 10% 58 9% 374 13% SE 34 4% 95 15% 552 19% SU 116 15% 55 9% 471 16% SO 166 21% 108 17% 347 12% MB 16 2% 16 3% 19 1% LE 68 9% 76 12% 493 17% OE 178 22% 124 20% 208 7% Total 798 100% 619 100% 2,880 100% Source: Traffic Lights PPMI (2018) and Smart City PMI (2020). F IGURE 38. P ROPORTION OF CONTROLLERS BY COMMUNICATION INFRASTRUCTURE Underground ducts 19% Aerial cabling No infrastructure 14% 67% Source: Traffic Lights PPMI (2018) and Smart City PMI (2020). 7.1.3.2 Connection Type Connection type identifies if there are copper cables, optical fiber or both implemented in CET’s communication infrastructure (Annex C11. Duct Network Map and Figure 37). Currently, 31.1% of all controllers are covered by some type of connection, 541 (12.6%) by copper, 174 (4%) only by fiber, and 622 (14.5%) by both. The majority (68.9%) cannot be connected to a central because there is no available cabling. 96 of 134 C - Diagnosis T ABLE 35. Q UANTITY OF CONTROLLERS BY CONNECTION TYPE GET Copper % Fiber % Mixed % No connection % CN 113 21% 24 14% 81 13% 427 14% NO 35 6% 42 24% 57 9% 376 13% SE 59 11% 22 13% 45 7% 555 19% SU 46 9% 21 12% 80 13% 495 17% SO 113 21% 12 7% 123 20% 373 13% MB 15 3% 0 0% 17 3% 19 1% LE 37 7% 29 17% 76 12% 495 17% OE 123 23% 24 14% 143 23% 220 7% Total 541 100% 174 100% 622 100% 2,960 100% Source: Traffic Lights PPMI (2018) and Smart City PMI (2020). F IGURE 39. P ROPORTION OF CONTROLLERS BY CONNECTION TYPE Fiber Copper 4% 13% Mixed 14% No connection 69% Source: Traffic Lights PPMI (2018) and Smart City PMI (2020). 7.1.4 Micro Mobility Diagnosis 7.1.4.1 Bikes Cycling Network As mentioned earlier, 2017 OD Survey identified a 45.2% growth in the number of trips made by bicycle in São Paulo since the last survey in 2007. This recent popularization of cycling is due to a mobility policy that gained momentum in 2007 with the creation of São Paulo’s Cycling System (City Law nº 14.266/2007) and its attribution to CET. Before the Cycling System creation, the SVMA (Municipal Secretary of Green and Environment) was responsible for all bicycle-related initiatives, which associated cycling with leisure activities in parks and green areas. When CET became responsible for São Paulo’s Cycling System, bikes started to be seen as a means of transport integrated to urban mobility, rather than just a leisure practice [33]. An example of this new perspective is the incentive to install bike racks near bus terminals and subway/train stations. 97 of 134 C - Diagnosis Ever since, São Paulo City Hall has invested in the consolidation of cycling infrastructure (Annex C12. Cycling Network Map and Figure 40), which currently has: ▪ A total of 681 km of roads with permanent cycling treatment, of which 649.4 km are painted/separated bike lanes and 31.6 km are bike routes. ▪ 117.4 km of temporary leisure cycling network (Sundays and national holidays, from 7 am to 4 pm). ▪ 72 indoor bike parking spaces with a total of 7,192 parking spots. ▪ 29 outdoor bike parking spaces with a total of 802 parking spots. At the end of 2019, São Paulo’s Cycling Plan was released, which establishes guidelines for the municipal cycling network growth and the improvement of existing infrastructure. In order to map all signalized intersections located at a road with bike lanes, georeferencing tools were used to associate the traffic lights geodatabase with São Paulo’s cycling network (Annex C6c. Traffic Light System Map – Cyclists). Currently, only a very small portion of signalized intersections have exclusive signal heads for cyclists (3.6% of total) and most are implemented at roads with bike lanes (94.3% of cyclist traffic lights). However, there are 1,538 traffic lights in the cycling network with no signaling for cyclists (26.1% of total). T ABLE 36. Q UANTITY OF SIGNALIZED INTERSECTIONS WITH SIGNAL HEADS FOR CYCLISTS Cyclist traffic light GET Yes % No % CN 83 39% 769 14% NO 22 10% 685 12% SE 17 8% 899 16% SU 22 10% 866 15% SO 11 5% 916 16% MB 1 0% 63 1% LE 23 11% 781 14% OE 32 15% 696 12% Total 211 100% 5,675 100% Source: Traffic Lights PPMI (2018). 98 of 134 C - Diagnosis F IGURE 40. C YCLING NETWORK MAP Source: Future Mobility for São Paulo. 99 of 134 C - Diagnosis F IGURE 41. T RAFFIC L IGHT SYSTEM MAP – CYCLISTS Source: Future Mobility for São Paulo. 100 of 134 C - Diagnosis F IGURE 42. P ROPORTION OF SIGNALIZED INTERSECTIONS WITH SIGNAL HEADS FOR CYCLISTS Yes 4% No 96% Source: Traffic Lights PPMI (2018). T ABLE 37. Q UANTITY OF SIGNALIZED INTERSECTIONS CONSIDERING THE CYCLING NETWORK No bike lane Bike lane Total No cyclist traffic light 4,137 1,538 5,675 Cyclist traffic light 12 199 211 Total 4,149 1,737 5,886 Source: Traffic Lights PPMI (2018) and SPTrans (2021). Cyclists According to a report published by SMUL (Municipal Secretary of Urbanism and Licensing), the 2017 OD Survey registered 59,500 cyclists and 220 thousand trips made by bicycles in São Paulo. Among the reasons listed by the survey, commuting to work (67.9%) is the main reason to for using a bicycle, followed by commuting to school (15.5%) and leisure activities (8.5%). Almost all trips are made only by bicycle (96.6%), without integration with other means of transport [18]. T ABLE 38. C YCLISTS PER FAMILY INCOME IN S ÃO PAULO Family Income Cyclists 2007 % Cyclists 2017 % Variation (%) Up to 2 MW 15,450 33.7 11,233 22.2 -27.3 From 2 up to 4 MW 19,736 43.1 20,041 39.6 1.5 From 4 up to 8 MW 9,018 19.7 11,086 21.9 22.9 From 8 up to 12 MW 542 1.2 4,094 8.1 655.4 More than 12 MW 1,093 2.4 4,102 8.1 275.3 Total 45,839 100 50,556 100 10.3 Source: PMSP, 2020. About the profile of cyclists in São Paulo, most remain among low-income groups – 61.9% of cyclists have a family income of up to 4 minimum wages. The bicycle is, historically, a mode used by low-income 101 of 134 C - Diagnosis population to reduce transport costs. However, comparisons between 2007 and 2017 indicate a change in the socioeconomic patterns of cyclists, with an increase of bicycle use by upper classes. During this period, the number of cyclists with family income above 12 minimum wages increased by 275.3%. At the same time, there was a 27.3% drop among low-income bicycle users (up to 2 minimum wages). Regarding other indicators, the vast majority of cyclists in São Paulo are men (89.5%), with an average age of 33.7 years, and have graduated from high school or have unfinished higher education (37.7%) (PMSP, 2020). Also, the opinion poll “Viver em São Paulo 2020” asked respondents who did not use bicycles what changes would motivate them to become cyclists, with safety being the most mentioned factor (32%) followed by the construction of more bicycle lanes (18%). Another relevant point of this survey is that 25% of respondents who are not cyclists said they would never use a bicycle to travel around São Paulo, regardless of the changes. To evaluate the flow of cyclists over time, counting data from three bicycle lanes in São Paulo were analyzed: Brigadeiro Faria Lima Avenue, located in an upscale region of São Paulo that concentrates corporate buildings and companies from the financial and technological sector; Vergueiro Street, a mixed use street (commercial and residential) that connects the south zone to São Paulo city center; and Dr. Gastão Vidigal Avenue, where CEAGESP (São Paulo Warehouse and General Storage Company) is located. Overall, Figure 43 shows that the number of bicycle lane users has been growing since 2018 and, even with negative peaks during Covid-19 Pandemic, in March/2021 the number of active cyclists registered was above the 2019 average. It is interesting, however, to note differences in bicycles volumes between the three bicycle lanes analyzed during the pandemic: while the number of cyclists remained stable on Dr. Gastão Vidigal Avenue, it decreased on Brigadeiro Faria Lima Avenue and increased on Vergueiro Street. F IGURE 43. C YCLIST COUNTING (BOTH DIRECTIONS ) ON BICYCLE LANES LOCATED AT BRIGADEIRO FARIA LIMA A VENUE , VERGUEIRO S TREET , AND DR . GASTÃO V IDIGAL AVENUE 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 Jun-16 Jun-17 Jun-18 Jun-19 Jun-20 Dec-15 Dec-16 Dec-17 Dec-18 Dec-19 Mar-20 Dec-20 Mar-16 Sep-16 Mar-17 Sep-17 Mar-18 Sep-18 Mar-19 Sep-19 Sep-20 Mar-21 Faria Lima R. Vergueiro Av. Gastão Vidigal Total Source: CET. Brigadeiro Faria Lima Avenue is an important business axis in São Paulo. Therefore, social distancing measures and closure of non-essential services strongly affected people circulation in the region, which would justify the reduction of cyclists in this period. In addition, renovation works in this bicycle lane 102 of 134 C - Diagnosis started in August/2020, as provided by the São Paulo Cycling Plan, which could have difficulted cycling in the region. In the case of Vergueiro Street, bicycle volume has grown steadily since the beginning of the pandemic, which could be a consequence of several factors. With temporary closing of stores and restaurants decreed by the State of São Paulo Government in March/2020, the use of delivery applications has become very popular as an alternative to local consumption. As it serves as a connection between the Paulista Avenue region – a place with a concentration of commerce and services – and middle-class residential neighborhoods in the south zone, a greater traffic of delivery cyclists could be the reason for this increase. Besides, the fear of public transport occupancy could also have led users of collective means to adopt the bicycle during the pandemic. 7.1.4.2 Pedestrians Walking represents 28.1% of all trips originated in São Paulo when considering the main mode of travel and it is the most common transportation modal identified by 2017 OD Survey. Based on the OD Survey database, these daily trips are usually short – with an average duration of 12 minutes – and 70% occur inside the same OD Zone15. In addition, going to school (48%) and to work (32.6%) are the most cited reasons for this mean of transport. Regarding profile, almost half of the population who travels by foot as a main mean of transportation is under 18 years old (45%), with a slight predominance of women (54.1%) over men (45.9%). When it comes to family income, most are from low-income groups – 73.3% belongs to families who earn up to four minimum wages. Upper classes (family income over 12 minimum wages) represent only 2% of people who travels exclusively by foot. In an urban environment, sidewalks are the most important spaces for pedestrian circulation. In São Paulo, owners of a building or property are responsible for the maintenance of the sidewalk adjacent to it – this includes citizens, private entities, and public institutions. However, according to São Paulo’s Mobility Plan (2015), the division of responsibility between private owners and municipal government does not guarantee a quality sidewalk system. In this context, the Mobility Plan adopts PEC (Emergency Plan for Sidewalks) as a public initiative to renovate sidewalks incorporating accessibility guidelines. The PEC or Emergency Plan for Sidewalks was created by Law nº 14.675/08 in 2008 to enable necessary civil works to renovate or build sidewalks in São Paulo that were not implemented according to the respective legislation, especially regarding accessibility and pedestrian security. Nonetheless, it was only in 2019 that São Paulo City Hall established the emergency routes covered by PEC (Decree nº 58.845/19), prioritizing sources of pedestrian generation – including places of public or private services – in synergy with public transportation (Annex C6b. Traffic Light System Map – Pedestrians and Figure 44). In September 2021, São Paulo City Hall stated [34] that over 1.6 million square meters of sidewalks had been renovated in the city as defined by 2019 PEC. Civil works were conducted in all regions of São Paulo to guarantee a levelled and obstacle-free route for pedestrians, including signaling, tactile paving, and 15 2017 OD Survey divides São Paulo Metropolitan Region in 517 Origin-Destination Zones. These zones are defined according to urban and socioeconomic homogeneity patterns, among other technical criteria, and are the smallest geographical unit for statistical representativeness of data collected. For this, the city of São Paulo is divided in 342 OD Zones. 103 of 134 C - Diagnosis ramps for accessibility. Therefore, routes listed by 2019 PEC were adopted in this study as an indicator of pedestrian prioritization, much like bike lanes for bicycles or bus corridors for buses. 104 of 134 C - Diagnosis F IGURE 44. T RAFFIC L IGHT SYSTEM MAP – PEDESTRIANS Source: Future Mobility for São Paulo. 105 of 134 C - Diagnosis Analyzing the relation between traffic lights and pedestrian prioritization (Annex C6b. Traffic Light System Map – Pedestrians and Figure 44), Table 39 and Figure 45 show that most traffic lights in São Paulo currently have a pedestrian signal head (61.3% of total). Of those 3,610 signalized intersections, 76.89% are actuated by push buttons (Table 40), which means 834 pedestrian traffic lights do not have push buttons to indicate demand. In addition, with georeferencing tools it was possible to associate the location of traffic lights and 2019 PEC routes as a way to map areas with a high concentration of pedestrians and no exclusive traffic signaling. Thus, Table 42 shows that 31% of the 3,401 traffic lights located at PEC routes do not have pedestrian signal heads. T ABLE 39. Q UANTITY OF SIGNALIZED INTERSECTIONS WITH SIGNAL HEADS FOR PEDESTRIANS Pedestrian traffic light GET Yes % No % CN 560 16% 292 13% NO 412 11% 295 13% SE 501 14% 415 18% SU 579 16% 309 14% SO 537 15% 390 17% MB 41 1% 23 1% LE 536 15% 268 12% OE 444 12% 284 12% Total 3,610 100% 2,276 100% Source: Traffic Lights PPMI (2018). F IGURE 45. P ROPORTION OF SIGNALIZED INTERSECTIONS WITH SIGNAL HEADS FOR PEDESTRIANS No 39% Yes 61% Source: Traffic Lights PPMI (2018). 106 of 134 C - Diagnosis T ABLE 40. Q UANTITY OF SIGNALIZED INTERSECTIONS CONSIDERING PEDESTRIAN SIGNAL HEADS AND PUSH BUTTONS No push button Push button Total No pedestrian traffic light 2,200 76 2,276 Pedestrian traffic light 834 2,776 3,610 Total 3,034 2,852 5,886 Source: Traffic Lights PPMI (2018). T ABLE 41. Q UANTITY OF SIGNALIZED INTERSECTIONS IN PEC ROUTES Emergency Plan for Sidewalks (PEC) GET Yes % No % CN 510 15% 342 14% NO 403 12% 304 12% SE 524 15% 392 16% SU 513 15% 375 15% SO 488 14% 439 18% MB 5 0% 59 2% LE 479 14% 325 13% OE 479 14% 249 10% Total 3,401 100% 2,485 100% Source: Traffic Lights PPMI (2018) and Decree nº 58.845/19 – Map of PEC routes (2019). F IGURE 46. P ROPORTION OF SIGNALIZED INTERSECTIONS IN PEC ROUTES No 42% Yes 58% Source: Traffic Lights PPMI (2018) and Decree nº 58.845/19 – Map of PEC routes (2019). 107 of 134 C - Diagnosis T ABLE 42. Q UANTITY OF SIGNALIZED INTERSECTIONS CONSIDERING THE PEDESTRIAN NETWORK No PEC route PEC route Total No pedestrian traffic light 1,221 1,055 2,276 Pedestrian traffic light 1,264 2,346 3,610 Total 2,485 3,401 5,886 Source: Traffic Lights PPMI (2018) and Decree nº 58.845/19 – Map of PEC routes (2019). 7.1.4.3 Bike and Scooter Sharing The first public bicycle sharing system in Brazil was “Pedala Rio”, inaugurated in 2008 in Rio de Janeiro and later restructured to the Bike Rio format. The system works with fixed stations, where the user unlocks the bicycle with an application and returns it at another station within a set period of time. The initiative – consolidated by the partnership between Rio de Janeiro City Hall, Itaú Unibanco and Serttel business group – spread to other Brazilian cities and in 2012 reached São Paulo under the name Bike Sampa. In 2017, the bicycle sharing system sponsored by Itaú Unibanco started to be operated by Tembici mobility startup. This operator switch was accompanied by a change in technology, with the adoption of solutions from the Canadian company PBSC Urban Solutions to improve service quality. In São Paulo, the renovation of Bike Sampa system took place in 2018 and currently there are 260 stations spread across strategic points in São Paulo. Note in Annex C12. Cycling Network Map and Figure 40 that the distribution of stations is limited to the expanded city center perimeter. Access to the system is done by purchasing packages in the application, ranging from single trips (BRL 2.99 for 15 minutes), leisure (BRL 20.00 for 48 hours) or a fixed monthly fee (BRL 29.90 for 4 45-minute trips a day or BRL 39.90 for 4 60-minute trips). Data from January 2019 registered an average of 7,000 trips on weekdays and 4,000 on weekends, which represents an increase of 500% compared to the total for January 2017 [35]. In the 2018 Bike Sampa reformulation, other improvements to the system were announced, such as the possibility of using Bilhete Único to unlock shared bicycles, and the creation of “Estações Bike” – peripheral terminals that allow 12-hour bicycle rentals for long journeys. However, no updated information was found on the progress of these initiatives. With a similar operation to Bike Sampa, CicloSampa is a bicycle sharing system with fixed stations that also operates in São Paulo. Inaugurated in 2013, CicloSampa currently has 12 stations located in the expanded city center and close to points of interest, such as parks and train or subway stations. The service is free for up to 30 minutes of use, with a fee of BRL 6.00 for every 30 minutes exceeded. In 2018, Yellow startup arrived in São Paulo with a proposal of dockless shared bicycles, a system that does not require the use of fixed stations. After an initial test phase with 500 bicycles in the city, the company consolidated its operation within a restricted area in neighborhoods of the expanded city center. The service started with a price of BRL 1.00 for 15 minutes of use. Also in 2018, Yellow expanded its market by sharing electric scooters in the Itaim Bibi neighborhood and, in the following year, sharing electric bicycles. The company's growth – supported by millionaire 108 of 134 C - Diagnosis contributions from international investors – paved the way for Yellow’s merger with Grin in January 2019. As a result, Grow Mobility Inc. was born. Grin was already working with electric scooter sharing in São Paulo in partnership with Rappi, a popular delivery startup in the city. Its merger with Yellow allowed the consolidation of Grow Mobility Inc. as one of the most relevant companies in this sector. Another electric scooter sharing company present in São Paulo is Lime, which started operating in the city in July 2019. However, as part of a global strategy for financial sustainability, the company ended its activities in Brazil in January 2020. The rapid spread of electric scooters in 2019 sparked a public debate between São Paulo City Hall and micro mobility companies regarding the regulation of this new mean of transport. Also in May 2019, São Paulo City Hall published Decree nº 58.750/2019, which established provisional rules for the management and circulation of electric scooters activated by digital platforms, and SMT Ordinance nº 69/2019, which defined the registration process of these companies together with the Municipal Secretary of Mobility and Transport. After that, in August, the regulation of electric scooter sharing services was made official by Decree nº 58.907/2019. Among all established measures, the following can be highlighted: maximum speed of 20 km/h; circulation permitted only on bicycle lanes and roads with a maximum speed of 40 km/h; ban on users under 18 years old; scooters can be parked only in specific places (approved in the company's SMT registration). In addition, non-registered companies that offer scooter sharing services will be fined and their equipment will be confiscated. In addition to the friction between companies and the City Hall Covid-19 Pandemic had a strong impact on São Paulo’s urban micro mobility sector. Uber, which in March 2020 had inaugurated its own electric scooter rental system in the city, decided to close this segment in July 2020. The scooters from Uber that had been removed from the streets because of the pandemic did not return to circulation. Similarly, the drop in daily commuting trips directly affected Grow Mobility Inc. sales, which were already showing sign of deceleration since January 2020. As a result, the company filed for judicial recovery in July 2020 with debts totaling approximately BRL 38 million. 7.1.5 Mobility and Territory Additional Inputs In addition to sources mentioned in Chapter 5 – Traffic Light Data Collection, there are some topics related to general aspects of mobility and territory that were incorporated as indicators in the traffic light characterization methodology (see item 7.2 of this report) for a comprehensive view of the city. Given the territorial dimension of São Paulo, these indicators were analyzed based on districts – administrative units of São Paulo City Hall, there are 96 districts in total. The aim of this item (7.1.5) is to briefly explain the consulted sources of information and its validity. 7.1.5.1 Quantity of Jobs The quantity of existing jobs in each district is an information published by São Paulo City Hall based on RAIS (Annual Social Information Registration). Every company in Brazil must send a RAIS report to the Federal Government of its condition – company information, number of employees, type of contracts, and so on – for monitoring and tax control. 109 of 134 C - Diagnosis To understand where jobs are concentrated and which areas of São Paulo attract more trips, the total number of jobs – including commerce, services, industry, and construction – was distributed according to districts. In addition, a job index published by Rede Nossa São Paulo was accounted for. It is a formal employment index, representing the number of jobs for ten habitants in working age (15 years or older). T ABLE 43. Q UANTITY OF JOBS PER DISTRICT – TOP 5 AND BOTTOM 5 Position District Quantity of jobs 1st Itaim Bibi 340,577 2nd Santo Amaro 184,877 3rd Vila Mariana 149,752 4th Pinheiros 149,725 5th Jardim Paulista 143,371 92nd Guaianazes 7,095 93rd Cidade Tiradentes 6,783 94th Iguatemi 6,568 95th Perus 6,174 96th Marsilac 163 Source: RAIS (2019). T ABLE 44. J OB INDEX PER DISTRICT – TOP 5 AND BOTTOM 5 Position District Job index 1st Barra Funda 59.24 2nd Sé 46.40 3rd Itaim Bibi 34.62 4th Bom Retiro 22.72 5th Bela Vista 21.69 92nd Brasilândia 0.47 93rd Tremembé 0.44 94th Anhanguera 0.40 95th Iguatemi 0.35 96th Cidade Tiradentes 0.24 Source: Mapa da Desigualdade (2019) – Rede Nossa São Paulo. 7.1.5.2 Land Uses Another aspect relevant to understand the territorial distribution in São Paulo is use of land, which characterizes city areas according to economic, cultural, and social activities that occur in a given space or building. São Paulo City Hall categorizes each block per predominant land use, there are 16 registered uses: (0) No information; (1) Low-income horizontal residential; (2) Medium/high-income horizontal residential; (3) Low-income vertical residential; (4) Medium/high-income vertical residential; (5) Trade and services; (6) Industry and warehouses; (7) Residential, trade and services; (8) Residential, industry and 110 of 134 C - Diagnosis warehouses; (9) Trade, services and industry; (10) Garages; (11) Public buildings; (12) Schools; (13) Vacant site; (14) Others; and (15) No predominance. For processing, each block had its area calculated in square meters and was grouped by category. Finally, for every district an area percentage was calculated for these groups – for example, residential use area per district divided by a district’s total area. T ABLE 45. G ROUPS FOR USE OF LAND Group Use of land (1) Low-income horizontal residential (2) Medium/high-income horizontal residential Residential (3) Low-income vertical residential (4) Medium/high-income vertical residential (5) Trade and services Trade and industry (6) Industry and warehouses (9) Trade, services and industry (7) Residential, trade and services Mixed use (8) Residential, industry and warehouses (10) Garages Facilities (11) Public buildings (12) Schools (0) No information (13) Vacant site No predominance (14) Others (15) No predominance Source: São Paulo City Hall (2016). 7.1.5.3 OD Survey Metrô’s Origin-Destination Survey (2017) was a recurrent source of information for this study, especially regarding modal split in the city of São Paulo. To spatially distribute this information, quantity of daily trips were allocated by district of origin and grouped by main mode of travel: T ABLE 46. G ROUPS FOR DAILY TRIPS BY MAIN MODE OF TRAVEL Group Main mode of travel Subway Train Public transport Bus Charter Transport School Bus/Van Walking Micro mobility Bike 111 of 134 C - Diagnosis Group Main mode of travel Car (driver) Car (passenger) Motorbike (driver) Private transport Motorbike (passenger) Conventional Taxi Unconventional Taxi Source: Metrô OD Survey (2017). In addition, since pedestrian demand was one indicator difficult to geolocate – public transport has corridors, cyclists have bike lanes, cars have road hierarchy, and so on –, an additional analysis was carried out. To identify regions that concentrate pedestrians, daily trips by foot (walking as main mode of travel) with the same origin and destination (intrazonal trips) were calculated in each OD Zone, and later aggregated as districts. T ABLE 47. Q UANTITY OF INTRAZONAL TRIPS BY FOOT – TOP 5 AND BOTTOM 5 Position District Intrazonal trips by foot 1st Grajaú 204,616 2nd Jardim São Luís 155,652 3rd Capão Redondo 151,176 4th Jardim Angela 144,583 5th Sacomã 138,740 92nd Alto de Pinheiros 13,124 93rd Vila Leopoldina 10,160 94th Pari 9,663 95th Jaguará 6,583 96th Marsilac 1,784 Source: Metrô OD Survey (2017). 7.1.5.4 Population Density Lastly, population distribution in São Paulo is an important factor to consider when talking about mobility. For this, demographic data was taken from IBGE’s last Census (2010) and aggregated into districts. Then, the total number of habitants was divided by area to achieve population density. São Paulo’s average population density in 2010 was 74.58 habitants/acre. T ABLE 48. P OPULATION DENSITY BY DISTRICT – TOP 5 AND BOTTOM 5 Position District Population density 1st Bela Vista 267.15 2nd República 247.74 3rd Cidade Ademar 222.23 4th Santa Cecília 214.66 112 of 134 C - Diagnosis Position District Population density 5th Sapopemba 210.76 92nd Socorro 29.29 93rd Barra Funda 25.68 94th Anhanguera 19.78 95th Parelheiros 8.55 96th Marsilac 0.41 Source: IBGE Census (2010). 7.2 METHODOLOGY FOR TRAFFIC LIGHTS CHARACTERIZATION The objective of this characterization will be to group the intersections per level of complexity operation regarding, as mentioned, the following dimensions: Dimension 1: Current available infrastructure Dimension 2: Infrastructure usage trends Dimension 3: Communication and connection to central traffic control The above will result in an intersection clustering, that in further steps, will give tools to infer which improvement measures or interventions for their benefit, fits better to each group or typology. With the information from selected indicators, the characterization was carried out by means of a series of excluding filters. Intersections are classified according to their main characteristics to identify the current level of complexity operation. Then, in later steps, these groups will be useful to infer the ideal control mode and type of prioritization that will be implemented at some point within the horizon of a 100% wireless solution. 7.2.1 Dimension 1: Current available infrastructure Grouping intersections based on infrastructure characteristics and in the area of influence of existing emergency facilities to infer the need for prioritization and their operation mode in this regard. This dimension will allow sizing the hardware characteristics and control strategies of the upgraded system. 7.2.1.1 Evaluation indicators used: The evaluation indicators used as input for this dimension are described below. • Road Hierarchy, in the following order from the largest section to the smallest one, regarding both directions at each intersection: o Highways o VTR o Structural Arterial o Arterial o Collector o Local o Pedestrian For this indicator, a closer look at the intersections grouped by CET’s Road Hierarchy (see item 7.1.1.1) showed that CET assigns hierarchy based on the role of traffic corridors in the city, not on road geometry nor volume of passing vehicles. This was most evident in the Arterial 113 of 134 C - Diagnosis classification, where roads with different physical characteristics received the same classification. Therefore, for the purpose of this study only, it was deemed necessary to create a new hierarchy to further characterize the traffic light system. For this, the Road Network classification from São Paulo’s Master Plan was crossed referenced with CET’s Road Hierarchy to create a new category, “Structural Arterial” – where Arterial roads (CET) coincided with Structural roads N1 and N2 (Master Plan). Structural N1 are roads used as a connection between the city of São Paulo and other cities in the country. Additionally, Structural N2 are roads not included in N1 that are used as a connection between cities in the Metropolitan Region of São Paulo and Structural roads N1. They were selected due to their characteristics as an extension of highways. • Intersections with dedicated bike lanes. • Intersection belonging to a public transport corridor: o BRT corridor o Bus corridor o Exclusive lane for buses o On exclusive infrastructure: This category refers to intersections that have some type of exclusive infrastructure for public transport such as bus lanes or bus corridors. • Intersections within a radius buffer of emergency facilities: o Military police stations: 500 meters radius o Hospitals: 250 meters radius o Fire stations: 750 meters radius Although theoretically all traffic signalized intersections in the city should be prioritized for emergency vehicles, because an emergency can occur anywhere. For prioritization purposes, the decision was made to assign this prioritization only to intersections near emergency facilities, because these areas are where the greatest number of emergency vehicles are concentrated. The radius of influence for the emergency facilities were calculated in a differentiated manner according to their type. This differentiation was estimated based on the number of facilities through a spatial analysis and the order of magnitude of emergency vehicles expected for each type. 7.2.1.2 Methodology description The above evaluation indicators will be grouped into the following clusters, based on the size and complexity of the intersections. In this way, Figure 1 presents the intersections denoted XL due to the fact that they involve intersections that are generated at the junction of freeways, structural arterials and arterial roads. Among the possible combinations, the infrastructure dedicated to public transportation was taken into account, as shown in Table 49. T ABLE 49. F IRST CLUSTER TYPOLOGY (XL-INTERSECTIONS ) Clusters Road hierarchy Public Transport Active mobility 1.1 VTR or Highways - - Arterials both directions Pedestrian 1.2 or structural arterial in BRT or bus corridor road/crossing at least one direction or Bike lane 114 of 134 C - Diagnosis Clusters Road hierarchy Public Transport Active mobility Arterials both directions 1.3 or structural arterial in BRT or bus corridor - at least one direction Arterials both directions 1.4 or structural arterial in Exclusive lane for buses - at least one direction Arterials both directions Pedestrian 1.5 or structural arterial in - road/crossing at least one direction or Bike lane Arterials both directions 1.6 or structural arterial in - - at least one direction Source: Future Mobility for São Paulo Consortium. The second category of clusters, refers to L-intersections, which are generated at the intersection of arterial roads with less complexity than the XL-intersections. In this category, in addition to taking into account the infrastructure dedicated to public transport, active modes such as pedestrians and bicycle lanes were included, as shown in Table 50. T ABLE 50. S ECOND C LUSTER TYPOLOGY (L-INTERSECTIONS ) Clusters Road hierarchy Public Transport Active mobility Pedestrian 2.1 Arterial On exclusive infrastructure road/crossing 2.2 Arterial On exclusive infrastructure Bike lane 2.3 Arterial BRT or bus corridor - 2.4 Arterial Exclusive lane for buses - Source: Future Mobility for São Paulo Consortium. In the third category are M intersections that include crossings between arterial roads without dedicated public transport infrastructure and with active mode prioritization, as well as intersections between collector and local roads with dedicated public transport infrastructure and active mode prioritization, as shown in Table 51. T ABLE 51. T HIRD C LUSTER TYPOLOGY (M- INTERSECTIONS ) Clusters Road hierarchy Public Transport Active mobility Pedestrian 3.1 Arterial - road/crossing 3.2 Arterial - Bike lane 3.3 Arterial - - Pedestrian 4.1 Collector On exclusive insfrastructure road/crossing 4.2 Collector and local On exclusive insfrastructure Bike lane 4.3 Collector both directions On exclusive insfrastructure Bike lane 115 of 134 C - Diagnosis 4.4 Collector and local BRT or bus corridor - 4.5 Collector both directions BRT or bus corridor - 4.6 Collector and local Exclusive lane for buses - 4.7 Collector both directions Exclusive lane for buses - Source: Future Mobility for São Paulo Consortium. The fourth category includes S-intersections, which correspond to crossings between collector and local roads without public transport infrastructure, but with prioritization of active modes and intersections between collector and local roads with the presence of bus lines, as shown in Table 52. T ABLE 52. F OURTH CLUSTER TYPOLOGY (S- INTERSECTIONS ) Clusters Road hierarchy Public Transport Active mobility Pedestrian 5.1 Collector - road/crossing 5.2 Collector and local - Bike lane 5.3 Collector both directions - Bike lane 5.4 Collector and local Bus line - 5.5 Collector both directions Bus line - Source: Future Mobility for São Paulo Consortium. The last category refers to XS-intersections, which are intersections between collector and local roads that do not have public transport routes or active mode prioritization and correspond to the intersections with the lowest level of complexity within all categories, as shown in Table 53. T ABLE 53. F IFTH C LUSTER TYPOLOGY (XS- INTERSECTIONS ) Clusters Road hierarchy Public Transport Active mobility 6.1 Collector and local - - 6.2 Collector both directions - - 6.3 Others on local roads - - Source: Future Mobility for São Paulo Consortium. Intersections within the radius of influence defined for emergency facilities will have this attribute across the board no matter in which cluster they were classified. 7.2.1.3 Results – Dimension 1 As a result of applying the methodology described above, in the Table 54 the results of each cluster are shown. Also, this result can be visualized in Annex C13. Dimension 1 Map (Figure 47) and its agglomeration by district in Annex C16 Districts Summary – Dimension 1 Map (Figure 48). T ABLE 54. RESULTS OF DIMENSION 1 Quantity of Clusters Description of joints intersections 1.1 VTR or Highways 12 116 of 134 C - Diagnosis Quantity of Clusters Description of joints intersections Arterial both directions or structural arterial + (BRT corridors or bus corridors) + 1.2 449 Pedestrians 1.3 Arterial both directions or structural arterial + (BRT corridors or bus corridors) 184 1.4 Arterial both directions or structural arterial + ExLB 321 1.5 Arterial both directions or structural arterial + (Pedestrian or Bike) 176 1.6 Arterial both directions or structural arterial 348 2.1 Arterial + Public transport exclusive infrastructure + Pedestrian 170 2.2 Arterial + Public transport exclusive infrastructure + Bike 220 2.3 Arterial + (BRT corridors or bus corridors) 192 2.4 Arterial + ExLB 378 3.1 Arterial + Pedestrians 201 3.2 Arterial + Bike 239 3.3 Others on arterial roads 560 4.1 Collector + Local + Public transport exclusive infrastructure+ Pedestrian 42 4.2 Collector + Local + Public transport exclusive infrastructure+ Bike 20 4.3 Collector both directions + Public transport exclusive infrastructure+ Bike 35 4.4 Collector + Local + (BRT corridors or bus corridors) 5 4.5 Collector both directions + (BRT corridors or bus corridors) 15 4.6 Collector + Local + ExLB 58 4.7 Collector both directions + ExLB 165 5.1 Collector + Pedestrian 230 5.2 Collector + Local + Bike 143 5.3 Collector + Bike 283 5.4 Collector + Local + Bus line 336 5.5 Collector both directions + Bus line 628 6.1 Collector + Local 152 6.2 Collector both directions 272 6.3 Others on local roads 52 Source: Future Mobility for São Paulo Consortium. 117 of 134 C - Diagnosis F IGURE 47. D IMENSION 1 MAP Source: Future Mobility for São Paulo. 118 of 134 C - Diagnosis F IGURE 48. D ISTRICTS SUMMARY – DIMENSION 1 MAP Source: Future Mobility for São Paulo. 119 of 134 C - Diagnosis The distribution of the results obtained in the dimension according to the numbering of the Table 54 is presented graphically as follows in Figure 49. F IGURE 49. P ROPORTION OF INTERSECTIONS IN DIMENSION 1 3.1 3.2 2.3 3% 4% 3% 1.3 6.2 5.1 2.2 3% 1.5 5% 5.3 4% 4% 3% 2.1 3% 6.3 5% 4.6 4.1 4.3 1% 4.7 1% 1% 1% 4.2 3% 0% 1.4 5% Other 4.5 5.4 9% 5.2 0% 1.1 6% 0% 2% 1.6 4.4 5.5 0% 6% 6.1 11% 2.4 3% 6% 1.2 3.3 8% 10% Source: Future Mobility for São Paulo Consortium. 7.2.2 Dimension 2: Infrastructure usage trends Grouping intersections based on users and predominant means of transportation, focusing on public collective transportation and active mobility or micro-mobility. This dimension will help us infer which areas of the city could be used for initial pilot testing as well as will make it possible to prioritize interventions. 7.2.2.1 Evaluation indicators used: The evaluation indicators used as input for this dimension are described below. • Jobs index. • Uses of land, were grouped as follows: a. Residential group b. Commercial and industrial group: Trade, services, industry and warehouses c. Mixed group: a + b grouped d. Facilities group: Garages/parking, public buildings/facilities and schools e. No predominance group: no information, vacant sites, others and no predominance • Trips per transport mean: a. Collective transport: Collective Public Transport: Subway, train and bus Private collective services: Charter and school vans/buses b. Active Modes: walking and bike means of transport c. Individual Transport: Private vehicles: Car and motorcycle Individual public transport: Taxi and unconventional taxi 120 of 134 C - Diagnosis • Special infrastructure for active modes: Intersections in a 50m radius buffer of sidewalks with high flow of pedestrians (according to the PEC - Emergency Plan for Sidewalks in São Paulo). • Pedestrian Intrazonal trips. 7.2.2.2 Methodology description Districts will be grouped in the following way: Cluster 1.1 (High16 jobs index or commercial and industrial use predominance) and (active mobility as a predominant transport mean or high16 pedestrian intrazonal trips) and (high16 index of planned PEC routes). Cluster 1.2 (High16 jobs index or commercial and industrial use predominance) and (collective transport as a predominant transport mean). Cluster 1.3 (High16 jobs index or commercial and industrial use predominance) and (individual transport as a predominant transport mean). Cluster 2.1 (High16 population density or residential use predominance) and (active mobility as a predominant transport mean or high16 pedestrian intra zonal trips) and (high16 index of planned PEC routes). Cluster 2.11 (High16 population density or residential use predominance) and (active mobility as a predominant transport mean or high16 pedestrian intra zonal trips). Cluster 2.2 (High16 population density or residential use predominance) and (collective transport as a predominant transport mean) and (high16 index of planned PEC routes). Cluster 2.21 (High16 population density or residential use predominance) and (collective transport as a predominant transport mean). Cluster 2.3 (High16 population density or residential use predominance) and (individual transport as a predominant transport mean). Cluster 3 Equipment use predominance. Cluster 4 All other districts. 7.2.2.3 Results – Dimension 2 The number of intersections resulting from the application of the above clusters are shown in Table 55. Results can be visualized in Annex C14. Dimension 2 Map (Figure 51) and its agglomeration by district in Annex C17. Districts Summary – Dimension 2 Map (Figure 52). T ABLE 55. RESULTS OF DIMENSION 2 Clusters Use of land Transport mean Index of planned PEC Quantity of intersections Jobs, Commercial or Active mobility or 1.1 High 1,023 Industrial intrazonal trips 16 Percentile 75 121 of 134 C - Diagnosis Clusters Use of land Transport mean Index of planned PEC Quantity of intersections Jobs, Commercial or 1.2 Collective - 22 Industrial Jobs, Commercial or 1.3 Individual - 1,374 Industrial Densely populated or Active mobility or 2.1 High 229 residential intrazonal trips Densely populated or Active mobility or 2.11 - 869 residential intrazonal trips Densely populated or 2.2 Collective High 330 residential Densely populated or 2.21 Collective - 801 residential Densely populated or 2.3 Individual - 1,065 residential 3 Equipment - - 33 4 No predominance - - 140 Source: Future Mobility for São Paulo Consortium. F IGURE 50. P ROPORTION OF DISTRICTS IN DIMENSION 2 3 1% 4 2% 1.1 2.3 17% 1.2 18% 0% 2.21 14% 1.3 23% 2.2 6% 2.1 2.11 4% 15% Source: Future Mobility for São Paulo Consortium. 122 of 134 C - Diagnosis F IGURE 51. D IMENSION 2 MAP Source: Future Mobility for São Paulo. 123 of 134 C - Diagnosis F IGURE 52. D ISTRICTS SUMMARY – DIMENSION 2 MAP Source: Future Mobility for São Paulo. 124 of 134 C - Diagnosis 7.2.3 Dimension 3: Communication and connection to central traffic control This step will be based on the current communication infrastructure available at each intersection and the possibility of connection to a central of the controller currently installed (hardware specifications). The results of this analysis will be used to complementing the cost analysis and the possibility of using 5G technology. 7.2.3.1 Evaluation indicators used: The evaluation indicators used as input for this dimension are described below. ▪ Existing CET communication infrastructure:  Underground  Aerial ▪ Existing connection type:  Optical fiber  Metallic  Mixed (both metallic and optical fiber)  No current connection ▪ Hardware specifications of existing controllers, according to CET:  Centralizable (controller can be connected to a central traffic control if there are communication means).  Non-centralized (controller cannot be connected to a central). 7.2.3.2 Methodology description The evaluation indicators will be grouped in the following way: Cluster 1.1 Centralizable controller and underground metallic wiring. Cluster 1.2 Centralizable controller and underground optical fiber wiring. Cluster 1.3 Centralizable controller and underground mixed wiring. Cluster 1.4 Centralizable controller and underground duct (no wiring or no information). Cluster 1.5 Centralizable controller and aerial wiring. Cluster 1.6 Centralizable controller and (no wiring or no information). Cluster 2.1 Non-centralized controller and underground metallic wiring. Cluster 2.2 Non-centralized controller and underground optical fiber wiring. Cluster 2.3 Non-centralized controller and underground mixed wiring. Cluster 2.4 Non-centralized controller and underground no wiring information. Cluster 2.5 Non-centralized controller and aerial wiring. Cluster 2.6 Non-centralized controller and (no wiring or no information). 7.2.3.3 Results – Dimension 3 Once the steps described above had been completed, we will proceed through spatial and/or matrix analysis to perform the consolidated characterization, the results are shown in Table 56. Results can also be visualized in Annex C15. Dimension 3 Map (Figure 54) and its agglomeration by district in Annex C18. Districts Summary – Dimension 3 Map (Figure 55). 125 of 134 C - Diagnosis T ABLE 56. RESULTS OF DIMENSION 3 Controller Quantity of Clusters Infrastructure Connection Result specifications intersections Able to be connected to a 1.1 Centralizable Underground Metallic 182 central system Optical Able to be connected to a 1.2 Centralizable Underground 81 fiber central system Able to be connected to a 1.3 Centralizable Underground Mixed 356 central system Able to be connected only 1.4 Centralizable Underground - with new wiring or cellular 25 network Able to be connected only 1.5 Centralizable Aerial Any with new wiring or cellular 458 network Able to be connected only 1.6 Centralizable - - with new wiring or cellular 1,022 network Able to be connected only 2.1 Non-centralized Underground Metallic 18 changing the controller Optical Able to be connected only 2.2 Non-centralized Underground 41 fiber changing the controller Able to be connected only 2.3 Non-centralized Underground Mixed 41 changing the controller Able to be connected only 2.4 Non-centralized Underground - 54 changing the controller Currently not able to be 2.5 Non-centralized Aerial Any 161 connected Currently not able to be 2.6 Non-centralized - - 3,447 connected Source: Future Mobility for São Paulo Consortium. 126 of 134 C - Diagnosis F IGURE 53. P ROPORTION OF INTERSECTIONS IN DIMENSION 3 Source: Future Mobility for São Paulo Consortium. 127 of 134 C - Diagnosis F IGURE 54. D IMENSION 3 MAP Source: Future Mobility for São Paulo. 128 of 134 C - Diagnosis F IGURE 55. D ISTRICTS SUMMARY – DIMENSION 3 MAP Source: Future Mobility for São Paulo. 129 of 134 C - Diagnosis 8 GENERAL CONCLUSIONS The urbanization process in Latin American countries was carried out in an unprecedent way, with significant growth rates in recent decades and high concentration of population living in urban areas. These high rates of urbanization growth pose major challenges for cities, especially for a megacity like São Paulo which is the center of one of the largest metropolitan regions in the world. São Paulo’s urban mobility network is extensive, being composed of 20 thousand kilometers of roads, 101 kilometers of subway lines, 271 kilometers of train tracks, and the largest bus network in Latin America. Historically, urban policies in the region prioritized motorized individual transport, however due to its constant population and geographical expansion – in addition to a strong socioeconomic and spatial segregation – a large portion of São Paulo habitants is dependent on public transportation. Overall, there are more than 11 million trips by collective means of transportation (41.3% of total), of which more than half is by bus. Active modes of transport are also significant, representing 29% of daily trips identified in the OD Survey (28% walking and 1% bicycle use). São Paulo is progressively adopting initiatives aligned with active and sustainable transportation guidelines, such as sidewalk renovation programs, bike sharing, and expansion of the cycling network. These trends – high demand for bus services, high quantity of vehicles in circulation, and high number of non-motorized trips –, combined with a complex transportation system, require technologies that support flexible and on-demand solutions for traffic control. Given the dimension of São Paulo’s Road network, its traffic light system is composed of 5,886 regular traffic lights – of those, 4,296 are controllers and 1,590 are dependent intersections – divided into 8 management areas (GETs) and associated to 8 traffic control and monitoring centrals (DCSs). This infrastructure is operated by CET, which hires private companies for maintenance and modernization services by public bidding. Since the modernization scope depends on budget availability, there is no way to plan a long-term system renovation. For instance, the last contract only covered corrective maintenance and was renewed until present time for lack of a new bidding process. Modernization inconsistency led to problems in infrastructure, especially lack of cable insulation – which leads to Priority 1 failures that multiply tenfold during the rainy season –, detection network disconnection – which causes all controllers to operate in fixed time –, and lack of connection to a traffic control central. Regarding CET’s communication infrastructure, more than half is aerial (51.4% of total length) and 67% of controllers are not covered by any type of communication infrastructure, whether they are underground or overhead. In addition, only 18.5% of all controllers are covered by optical fiber. Another major issue – not directly related to infrastructure modernization – is failures caused by vandalism, which mainly comprises of theft of metal cables, theft of controller cards and breaking fiber optic cable. The IDOM study [28] had access to failure reports and concluded that theft and vandalism of traffic light equipment represent 25% of all occurrences and it demands 47% of maintenance resources. Furthermore, 23% of failures occur due to power shortage or equipment wear. These three reasons require 90% of the time used by maintenance while the rest 10% resources are used to resolve the other 50% of failures which occur due to uncontrollable factors. Moreover, São Paulo’s traffic lights don’t have a stable power supply system which could lead to accidents. Old installations and material owe to a lot of power faults in traffic light system. 130 of 134 C - Diagnosis The IDOM report [28] also pointed out the drawbacks in the “Control devices” of the entire control loop architecture. The other limitation is in the control strategy. The Control Strategy forms the center of the control loop and its relevance and efficiency determine primarily the efficiency of the overall control system. Figure 56 below illustrates the basic elements of a control loop. F IGURE 56. T HE CONTROL LOOP As for centralization, CET has adopted Siemen’s SCOOT system as the standard for São Paulo, as well as UTMC (type 2) and NTCIP for communication protocols between controller and traffic central. However, there are only 232 centralized controllers in the city and none operating in real time with SCOOT. To summarize, all traffic lights operate in fixed time – even the centralized ones – and the majority of controllers cannot be centralized due to limitations of the model (68% of total). Lastly, with increasing public transport demand and active mobility trends, it is essential for the traffic control to consider their priorities at signalized intersections. According to SPTrans, the current bus fleet does not have any kind of OBU (On Board Units) for traffic light prioritization, and the AVL sends GPS signals for itinerary monitoring only. Even though most intersections in São Paulo (4,990 out of 5,886) are located at a public transport corridor (bus line, bus corridor or exclusive lane for buses), SPTrans does not have plans for traffic light PuT prioritization as the current technology hinders any kind of implementation17. In addition, emergency vehicle prioritization is not adopted in the city. Considering all the problems mentioned above, the traffic light characterization detailed in Chapter 7 will be a basis for different solutions to be proposed in the next steps of this study (summary of all dimensions in Annex C19. Clustering GET’s Map and Figure 57). On top of this, telecommunication regulation specified in Chapter 3 should guide future recommendations for 5G employment in São Paulo. The goal of this study is to start from a basic upgrade of the current situation and reach a state-of-the-art solution (and further) to a zero human-machine interface control scenario – with the use of artificial intelligence and 5G. 17 Information provided during an internal meeting with SPTrans in August 2021. 131 of 134 C - Diagnosis F IGURE 57. C LUSTERING GET’S MAP Source: Future Mobility for São Paulo. 132 of 134 C - Diagnosis 9 BIBLIOGRAPHY [1] “Transportation Research Thesaurus.” https://trt.trb.org/search?term=crashes (accessed Oct. 22, 2021). [2] ABNT, “NBR 10697:2020 - Pesquisa de sinistros de trânsito - Terminologia,” 2020. [3] SWARCO, “Mobility Trends.” . [4] DPS Telecom, “Open Protocols Vs. Proprietary Protocols.” . [5] ManageEngine OpManager, “Network Protocols | Types of Networking Protocols.” . [6] Dataprom, “Soluções inteligentes de mobilidade,” 2019. [7] United NationsDepartment of Economic and Social Affairs, “World Urbanization Prospects,” 2019. [Online]. Available: https://population.un.org/wup/Publications/Files/WUP2018-Report.pdf. [8] EMPLASA, “Plano de Desenvolvimento Urbano Integrado - RMSP: diagnóstico final,” 2019. [Online]. Available: http://multimidia.pdui.sp.gov.br/rmsp/docs_pdui/rmsp_docs_pdui_0018_diagnostico_final.pdf. [9] Anatel, “Anatel aprova o edital do leilão do 5G.” https://www.gov.br/anatel/pt-br/assuntos/noticias/anatel- aprova-o-edital-do-leilao-do-5g (accessed Sep. 26, 2021). [10] Governo do Brasil, “Leilão do 5G confirma expectativas e arrecada R$ 47,2 bilhões — Português (Brasil),” Série Especial 5G, 2021. https://www.gov.br/pt-br/noticias/transito-e-transportes/2021/11/leilao-do-5g- confirma-expectativas-e-arrecada-r-47-2-bilhoes (accessed Nov. 25, 2021). [11] ABRINTEL, “História da Abrintel.” https://abrintel.org.br/conheca-a-abrintel/ (accessed Sep. 10, 2021). [12] B. do Amaral, “Compartilhamento de rede 3G entre Claro e Vivo é aprovado na Anatel,” TELETIME News, 2021. https://teletime.com.br/19/08/2021/compartilhamento-de-rede-3g-entre-claro-e-vivo-e-aprovado- na-anatel/ (accessed Sep. 10, 2021). [13] B. do Amaral, “TIM e Vivo firmam acordo de compartilhamento de rede 2G, 3G e 4G,” TELETIME News, 2019. https://teletime.com.br/19/12/2019/tim-e-vivo-firmam-acordo-de-compartilhamento-de-rede-2g-3g-e- 4g/ (accessed Sep. 10, 2021). [14] B. do Amaral, “TIM testa desligamento da rede 2G em 50 cidades,” TELETIME News, 2021. https://teletime.com.br/10/02/2021/tim-testa-desligamento-da-rede-2g-em-15-cidades/ (accessed Sep. 10, 2021). [15] B. do Amaral, “Rede 4G compartilhada entre Vivo e TIM deverá dobrar até final do semestre,” TELETIME News, 2021. https://teletime.com.br/24/02/2021/rede-4g-compartilhada-entre-vivo-e-tim-devera-dobrar- ate-final-do-semestre/ (accessed Sep. 10, 2021). [16] Anatel, “Antenas nos Municípios.” https://www.gov.br/anatel/pt-br/dados/infraestrutura/antenas-nos- municipios (accessed Sep. 20, 2021). [17] PMSP, “O que mudou na mobilidade no município de São Paulo entre 2007 e 2017?,” 2019. [Online]. Available: https://www.prefeitura.sp.gov.br/cidade/secretarias/upload/Informes_Urbanos/40_IU_OD_2019_final.pd f. [18] PMSP, “O uso da bicicleta na cidade de São Paulo OD 2007 e 2017,” 2020. [Online]. Available: https://www.prefeitura.sp.gov.br/cidade/secretarias/upload/Informes_Urbanos/42_IU_ciclistas_2020_fin al.pdf. [19] Rede Nossa São Paulo & IBOPE Inteligência, “Pesquisa de Opinião Pública Viver em São Paulo: Mobilidade Urbana.” p. 257, 2020, [Online]. Available: https://www.nossasaopaulo.org.br/wp- content/uploads/2020/11/ViverEmSP-MobilidadeUrbana-2020-completa.pdf. [20] Rede Nossa São Paulo & IBOPE Inteligência, “Pesquisa de Opinião Pública Viver em São Paulo: Mobilidade Urbana,” 2019. https://www.nossasaopaulo.org.br/wp- 133 of 134 C - Diagnosis content/uploads/2019/09/Pesquisa_ViverEmSP_MobilidadeUrbana_completa_2019.pdf. [21] Metrô, “Relatório Síntese - Pesquisa Origem Destino 2017,” 2019. [Online]. Available: http://www.metro.sp.gov.br/pesquisa-od/. [22] ITDP (Institute for Transportation and Development Policy) and WRI Brasil, “Onde estão as pessoas e o transporte na cidade de São Paulo,” 2016. [Online]. Available: http://itdpbrasil.org.br/wp- content/uploads/2017/02/2017-02-ITDP-PNT-SP.pdf. [23] DENATRAN, “Estatísticas - Frota de Veículos.” https://www.gov.br/infraestrutura/pt- br/assuntos/transito/conteudo-denatran/estatisticas-frota-de-veiculos-denatran (accessed Jun. 06, 2021). [24] CETESB (Companhia Ambiental do Estado de São Paulo), “Vehicular emissions in São Paulo state,” 2019. [Online]. Available: https://cetesb.sp.gov.br/veicular/relatorios-e-publicacoes/. [25] I. (Instituto de E. e M. Ambiente), “Emissões Atmosféricas (Gráficos),” Inventário de Emissões Atmosféricas do Transporte Rodoviário de Passageiros no Município de São Paulo, 2017. http://emissoes.energiaeambiente.org.br/graficos (accessed Oct. 15, 2021). [26] CET, “Acidentes de Trânsito - Relatório Atual,” 2019. [Online]. Available: http://www.cetsp.com.br/media/1058619/2019.pdf. [27] SP Parcerias S.A., “Carta Anual de Governança Corporativa,” 2019. [Online]. Available: https://www.spparcerias.com.br/sites/default/files/2019-03/CARTA ANUAL DE GOVERNANÇA CORPORATIVA.pdf. [28] IDOM & World Bank, “D1 - Status Quo Report,” Consultancy Services for alternatives analysis and other studies for traffic lights systems in São Paulo, 2019. . [29] D. H. Clemente, “Estratégia e Estrutura em Empresas de Mobilidade Urbana : o Caso da Companhia de Engenharia de Tráfego de São Paulo ( CET ),” Escola Politécnica da Universidade de São Paulo, 2017. [30] P. De Angelo, “Relatório sobre a implantação do primeiro controlador com Protocolo UTMC tipo 2 na Cidade de São Paulo,” 2014. [Online]. Available: http://www.cetsp.com.br/media/293419/nt_233.pdf. [31] L. M. Vilanova, “O Controle de Semáforos em Tempo Real - A Experiência de São Paulo,” 2005. [Online]. Available: http://www.cetsp.com.br/media/65265/bt38- o controle de semaforos em tempo real.pdf. [32] G. P. Kazukas, “Smart cities: uso de aplicativos de tecnologia para o planejamento urbano na cidade de São Paulo,” Universidade de São Paulo, 2018. [33] M. Harkot, “6 anos após junho de 2013 e das ciclovias e ciclofaixas, como está a mobilidade em SP?,” LabCidade, 2019. http://www.labcidade.fau.usp.br/6-anos-apos-junho-2013-e-da-implementacao-das- ciclovias-e-ciclofaixas-como-esta-a-mobilidade-em-sp/ (accessed Jul. 22, 2021). [34] PMSP, “Prefeitura disponibiliza mapa online das calçadas requalificadas na cidade,” 2021. https://www.prefeitura.sp.gov.br/cidade/secretarias/licenciamento/noticias/?p=318047 (accessed Sep. 17, 2021). [35] R. C. Rabello, “Sistema Público de Bicicletas Compartilhadas,” Universidade de São Paulo, 2019. 134 of 134 The information is provided by: Selection #1265335