An average of 6.5 million crashes are reported to the police every year in the United States. Safety is significantly important considering the rapid increase on traffic volume on American roads. This thesis describes the development of a safety model whose primary objective is to capture the benefits of Intelligent Transportation Systems (ITS) on safety. The specific ITS component that is examined in more detail is traffic signal coordination. The model was tested in a micro-simulation environment using INTEGRATION traffic simulation model as well as in a field data evaluation. The General Estimates System (GES) database was chosen as the primary national database to extract accident data. These data were used for the development of the statistical foundation for the safety model. Crash rates were produced using extracted crash frequencies and annual vehicle miles traveled figures from the Highway Statistics (FHWA, 1997). Regression analysis was performed to predict the behavior of several crash types, as they were associated with a variety of variables, for example the facility speed limit and time the crash occurred. The model was developed in FORTRAN code that estimates the accident risk of a facility based on its free-speed. Two methods were used to test the model: 1. field data from the city of Phoenix, Arizona were used in a GPS (Global Positioning Systems) floating car that tracked the accident risk on a second by second basis. Before and after signal coordination scenarios were tested thus yielding a result that the accident risk is less in the after scenario. 2. the model was then tested in a micro-simulation environment using the INTEGRATION traffic model. A hypothetical network, as well as the Scottsdale/Rural road corridor in Phoenix were used. The sensitivity analysis of before and after signal coordination scenarios indicated that after the signals were coordinated, the crash risk was lower, thus proving that the model could capture the benefits of this ITS component. Reducing the number of crashes is an important aspect of improving safety. Traffic signal coordination smoothens traffic on a facility and reduces its potential accident risk by producing less vehicle-to vehicle interactions. Also, traffic signal control increases the free-speed of a facility. The advantage of this safety model is the fact that it can be used to capture a variety of ITS technologies and not only signal coordination that is examined in more detail in this thesis. / Master of Science

Orientador: Professor Dr. João Manoel Losada Moreira / Esta pesquisa tem como objetivo fundamental contribuir para a supressão dos acidentes de trabalho fatais na cadeia produtiva de energia elétrica discutindo o problema em três frentes: a) estender o escopo dos bancos de dados de acidentes de trabalho fatais para toda a cadeia produtiva de energia elétrica incluindo etapas de empresas terceiras; b) incluir nos bancos de dados sobre acidentes de trabalho fatais informações que permitam avaliar as condições em que ocorreram os acidentes e formular soluções para evitá-los no futuro; c) monitorar e quantificar por meio de redes Bayesianas o nível da cultura de segurança das empresas da cadeia de energia elétrica admitindo-se uma correlação inversa entre o nível de cultura de segurança e o número de acidentes fatais nessas empresas. O setor de energia elétrica é normalmente representado pelas etapas geração, transmissão e distribuição de eletricidade. Contudo o setor elétrico, considerando toda sua cadeia produtiva é muito maior e foram identificadas 15 etapas desde a prospecção de energia primária até o descomissionamento das instalações após sua vida útil. Em relação aos bancos de dados de acidentes de trabalho, no Brasil há três bancos de dados que cobrem apenas as etapas de geração, transmissão e distribuição de energia elétrica. Foi encontrado que eles não apresentam dados completos que permitam a rastreabilidade das causas raízes dos acidentes e identifiquem as possíveis soluções para evitá-los no futuro. Por exemplo, entre 2009 e 2015 morreram 69 trabalhadores próprios e 380 terceirizados na etapa de distribuição sem maiores detalhes sobre as condições em que esses acidentes ocorreram. A inclusão das empresas terceiras na contabilidade das estatísticas de acidentes de trabalho pode ser um fator indutor de melhoria nas condições de trabalho de todo o setor de energia. O poder econômico das empresas líderes pode exigir de seus parceiros um comportamento mais adequado quanto à segurança do trabalho e supressão dos acidentes fatais. Foi proposta uma estrutura mínima de banco de dados de acidentes de trabalho no setor elétrico com informações que permitam entre outros objetivos a rastreabilidade e posterior correção dos eventos acidentais. Entre esses itens incluem-se o número de registro do acidente, nome, tipo e área de atuação da empresa e pareceres do sistema de acompanhamento da segurança do trabalho sobre acidente, por exemplo, CIPA, médico do trabalho, sindicato e supervisor técnico. A manutenção de um nível de cultura de segurança em uma empresa é importante para a supressão de acidentes de trabalho. Um modelo de monitoração da cultura de segurança baseado em redes bayesianas com 15 diferentes variáveis foi desenvolvido. Estas variáveis consideram desde o engajamento da alta administração com a segurança até treinamento de trabalhadores na utilização de equipamentos de proteção individual. Este modelo quantitativo conseguiu capturar diferenças de nível de cultura de segurança de 15 diferentes empresas de distribuição de eletricidade e mostrar uma clara correlação inversa com o número de acidentes fatais. / Tese ( doutorado)- Universidade Federal do ABC. Programa de Pós-Graduação em Energia, 2017. / This research aims at contributing to the suppression of fatal work accidents in the electricity sector, discussing the problem on three fronts: a) extend the scope of the fatal work accident data banks to the entire electrical energy production chain including services of third-party companies; b) include in the fatal accidents databases information to assess the conditions under which accidents occurred and formulate solutions to avoid them in the future; c) to monitor and quantify, through Bayesian networks, the level of safety culture of the companies in the electricity sector, assuming an inverse correlation between the level of safety culture and the number of fatal accidents in these companies. The electricity sector is usually represented by the stages of generation, transmission and distribution of electricity. However, the electrical sector, considering its entire production chain, is much larger and 15 stages have been identified from primary energy prospecting to the decommissioning of facilities after their useful life. Regarding work accident databases, in Brazil there are three databases covering only the generation, transmission and distribution stages of the electricity sector. It was found that they do not present complete data to trace the root causes of accidents and identify possible solutions to avoid them in the future. For example, between 2009 and 2015, 69 self-employed workers and 380 outsourced workers died in the distribution stage without further details of the conditions under which these accidents occurred. The inclusion of thirdparty companies in the accounting of work-accident statistics can be a factor in improving the working conditions of the entire energy sector. The economic power of leading companies can demand from their partners better behavior in terms of work safety and suppression of fatal accidents. It was proposed a minimum database structure of work accidents in the electric sector with information that allows, among other objectives, the traceability and subsequent correction of accidental events. These items include the accident record number, name, type and area of work of the company, and opinions of the accident work safety monitoring system, including those from the CIPA, occupational physician, union and technical supervisor. Maintaining a level of safety culture in a company is important for the suppression of occupational accidents. A safety culture monitoring model based on Bayesian networks with 15 different variables was developed. These variables range from senior management engagement with work safety to workers training in the use of personal protective equipment. This quantitative model was able to capture differences in the safety culture level of 15 different electricity distribution companies and show a clear inverse correlation with the number of fatal accidents.