Advanced Driver Assistance Systems (ADAS) are technologies that automate, facilitate, and improve the vehicle’s systems. Indeed, these systems directly interfere with braking, acceleration, and drivability of driving operations. These technologies directly reflect the psychology behind driving a vehicle, which can have an automation level that varies from fully manual (Level 0) to fully autonomous (Level 5). Even though ADAS systems provide safer driving, it is still a challenge to understand the complexity of human factors that influence and interact with these technologies. Also, there has been limited exploration of the correlation between the physical and cognitive reactions of the drivers and the characteristics of Brazilian roads and traffic. Objective: This work aims to analyze and characterize the human factors related to driver behavior while driving a vehicle with level 2 of automation, regarding the Lane Keeping Assist System (LKAS) and Forward Collision Warning System (FCWS), on roads in the Distrito Federal (DF).
Method: The driver and vehicle instrumentation were developed to conduct the study. The driver instrumentation was designed to be related to five physiological signals: heart rate, breathing, galvanic skin response, muscle activity, and brain activity. The study used tests in real traffic situations and virtual simulations to validate the designed data acquisition circuit. The driver behavior questionnaire was applied to select a driver group among the participants and identify the individual’s characteristics and relevant behavior trends. Results: In general, the results showed that drivers tend to rate themselves quite positively, although they admit to being prone to committing traffic errors and violations.
Furthermore, it was noted that the driver and vehicle interaction varies between participants, suggesting that behavior factors could directly affect the physiological response while driving. Thus, the case study showed that the previous driver’s experience with the vehicle’s systems technologies could impact the physiological response. Notwithstanding, the results showed the complexity inherent to emotional and cognitive states related to driving, as aspects such as velocity, the capacity to maintain the velocity, and the car’s acceleration and deceleration patterns could be associated with the driver’s characteristics. Conclusions: The research proved helpful in establishing a way to identify specific patterns that characterize the driver in the vehicle-driver environment. Moreover, it presented a preliminary evaluation of physiological responses relating to safety systems activated in the Brazilian traffic scenario, playing a role in understanding human factors in using driver assistance systems.

This study addresses short-term traffic forecasting using a Graph-CNN-LSTM-based model, which integrates spatial and temporal information to improve prediction accuracy. The data were collected by the Department of Highways of the Federal District (DER-DF) and include traffic volume records at 15-minute intervals from March to August 2024. During preprocessing, steps such as normalization, outlier treatment, and the construction of a dynamic adjacency matrix were performed, allowing for the modeling of interdependencies between road segments. The predictive model was evaluated using metrics such as Mean Absolute Error (MAE) and Root Mean Square Error (RMSE), demonstrating that prediction performance decreases as the forecast horizon increases. The average MAE values across all road segments were 13.70 (15 min), 14.30 (30 min), 19.39 (45 min), and 21.78 (60 min). The results indicate that high-traffic periods show greater variability in prediction errors, whereas lower-traffic hours allow for more accurate forecasts. The analysis of segment correlations revealed strong spatial interdependencies in traffic patterns, highlighting the importance of graph-based modeling. This study contributes to the enhancement of urban mobility management by providing a predictive model that can assist in strategic decision-making to optimize traffic flow and reduce congestion.

This study proposes the development of a parametric model to predict the costs of constructing or expanding a key element of airport infrastructure, notably the runway. Initially, the strategic importance of airports for economic and social development is highlighted, especially in the Brazilian context, where private management has become increasingly relevant. Additionally, the uniqueness of this word is emphasized, as it addresses the complexities of airport infrastructure in a specialized manner, filling a gap in the literature due to the need for specificity in studies in this field combined with the confidential nature of the costs involved. Furthermore, the importance of accurate cost estimates is discussed, emphasizing their influence on managerial decisions and the viability of airport infrastructure projects. For the development of the model, the parametric method was chosen due to its comparative advantages over other methodologies, particularly its potential simplicity and transparency. Using multiple linear regression technique, models are developed considering variables such as runway area, pavement classification, and airport location. The costs associated with runway construction and expansion were obtained from data provided by the Brazilian airport management state agency (INFRAERO) and budgets from Technical, Economic, and Environmental Feasibility Studies for the Brazilian airport concessions program. It is worth noting that these data are often unconsolidated, incomplete, or not applicable, requiring extensive and careful analysis in the process of data treatment and information selection. The study results demonstrate the effectiveness and feasibility of the proposed models in predicting the costs involved in the construction or expansion of runways, with coefficients of determination (R²) above 0.9 for some cases. However, some specific services, such as earthworks and signaling, faced significant challenges in their modeling, mainly due to the limitation of available data. The contribution of this study to the understanding and improvement of airport infrastructure management and development is significant.

This dissertation presents an environmental and cost assessment of the electrification of the Brazilian Air Force’s (FAB) fleet of light-duty vehicles. The scope of the study included Internal Combustion Engine Vehicles (ICEV), Hybrid Electric Vehicles (HEV), and Battery Electric Vehicles (BEV), using ethanol or gasoline. In this context, the Life Cycle Assessment (LCA) methodology was applied, according to the NBR ISO 14040 and NBR ISO 14044 standards, to quantify CO2eq emissions during the production and use phases of the vehicles. Additionally, an economic analysis was carried out, considering the acquisition, maintenance, and operational costs of the vehicles. An inventory model for emissions calculations was created for vehicles in use, based on the year and model data of each one, as well as for future acquisitions, depending on the chosen technology. The findings suggests that ethanol plays a key role in reducing emissions, and for sedans, combined with hybrid technology, the most sustainable results are achieved. For compacts, due to the lack of hybrid models, ethanol-powered ICEV are the most sustainable in most regions of the country, except in the Northeast, where, due to its clean energy matrix, BEV achieved better environmental performance over a 15-year period. The acquisition costs of electric vehicle (EV) remain very high compared to ICEV, suggesting that specific incentive policies for public institutions are necessary to make their adoption feasible. The study aims to support strategic decisions regarding fleet renewal through an integrated approach that considers both environmental and financial benefits, contributing to the sustainability and operational efficiency of the Air Force.

The study investigates the application of the iRAP (International Road Assessment Programme) methodology in a highway concession project in Paraná, Brazil, aiming to evaluate the effectiveness of road safety countermeasures and their cost-benefit relationship. The concession of highways to private initiatives has been adopted by the public sector as a solution to improve management, maintenance, and road safety, especially in the context of budget constraints. The iRAP methodology, recognized for its rigor in risk classification and intervention recommendations, was applied to a 500 km stretch of highways, selected based on criteria such as logistical relevance, traffic volume, and accident history. The analysis consisted of coding infrastructure characteristics to generate star ratings, allowing for risk prediction and the projection of cost-benefit scenarios for proposed measures. Using the Pearson Chi-Square test, significant correlations were identified between high-risk areas classified by iRAP and accident frequency. The results indicate that incidents occur more frequently in high-risk areas, and the suggested countermeasures, such as physical barriers, improved signage, and shoulder adjustments, are effective in reducing both the frequency and severity of accidents. These measures align with the concept of "forgiving roads," promoting significant improvements in road safety and mitigating risks of head-on collisions and pedestrian accidents. Financial analysis demonstrated that investments in road safety, as proposed by the iRAP methodology, generate substantial long-term economic benefits, outweighing the intervention costs. Furthermore, the research highlights the importance of data quality, as the effectiveness of countermeasures directly depends on the accuracy of information regarding road conditions and accident history. Thus, the study proposes a more assertive economic modeling of concession contracts and the incorporation of safety criteria in Road Exploration Programs (PER), enabling a replicable and sustainable model for infrastructure projects in Brazil.

The way energy is generated has changed dramatically with the advent of electricity and fossil fuels. In the 18th century, with the beginning of the industrial age, the concentration of atmospheric CO2 rose by 50%, being today 420 ppm (parts per million) corresponding to 150% of the value in 1750. Aiming to reduce the GHG emissions by road transport, several actions have been discussed to promote the use of more efficient and cleaner vehicles worldwide. Consequently, there has been an increase in demand for more efficient vehicles including hybrids, electrics and low-emission fuels such as ethanol, hydrogen and biodiesel. This study evaluates the impact on the Brazilian energy matrix, thus taking into account the growth of EVs in Brazil through the methodology of scenarios. It is observed that for the scenario called pessimistic in which the country grows only 1% per year, the current investment in energy would satisfy the future energy demands, including the increase in the number of electric cars. Nevertheless, for other growth scenarios, such as the most likely scenario (growth of 4% per year), Brazil would have to invest an average of R$ 12 billion reais just in new energy generating plants.

The free-flow system toll collection is already a reality in Brazil, having been implemented in part of the stretch concession called RioSP, and has the potential to generate major changes in tariff policies in road infrastructures. However, an important factor to be considered is the trend towards an increase in user defaults, as there is no physical barrier to prevent passage without paying the due toll rate. Thus, the objective of the study was to evaluate the toll rate impacts of the implementation of the free-flow system in a case study of the BR-101/RJ/SP. In this context, a systematic review of the literature was carried out, identifying the characteristics, advantages and disadvantages, the advances achieved by Brazil and the challenges to be overcome in order to expand the use of the free-flow system in the country. To assess the toll rate impacts on the cash flow of the BR-101/RJ/SP, two free-flow implementation scenarios were simulated: one just replacing the toll plazas provided for in the original contract; another proposing a pure free-flow model with charging per homogeneous segment at 17 charging points. To analyze the feasibility of the simulated scenarios, the concept of allowable non-payment was used, for which the economic-financial balance of the concession occurs with the kilometer toll rate and the project's original internal rate of return. The BR-101/RJ/SP concession’s Feasibility Studies were used to support the analysis of the free-flow implementation in part of the concession section. The momentum transport and billing system costs were calculated for the proposed model. As a result, on the one hand, it was found that the implementation of free-flow billing gateways at the same billing points originally studied is advantageous, resulting in a toll rate reduction of 14% or allowing margin to absorb up to 13.5% of frustration from annual revenue due to defaults. On the other hand, the simulation of charging per homogeneous segment on the highway under study, which increases 14 collection points, despite achieving the objective of tariff justice, implied a 24% increase in the kilometer toll rate in the base scenario.

In intelligent transportation, the great challenge is to achieve an efficient transportation system for the transport of people and goods. With the emergence of Autonomous Vehicles (AVs) new contextual and operational challenges bring the need for a coordinated development between the actors involved in the deployment of these modern technologies and the components necessary for the joint operation of AVs, pedestrians, infrastructure, and technological devices with the ability to communicate with each other efficiently and safely. The advantages and benefits arising from the operation of AVs on highways, in mixed or segregated traffic, will not be possible without the development of a quality and intelligent road infrastructure. In face of this new reality, the need arises to identify which components are necessary for the implementation of an intelligent infrastructure capable of receiving the AVs and meet their incorporation into the transportation system. In this sense, this dissertation aims to identify the components that characterize an intelligent highway in the Brazilian context and classify them according to their degree of importance. Based on a systematic literature review, five main methodologies were identified to classify intelligent highways: i) road infrastructure support levels for automated driving; ii) smart road classification; iii) classification of readiness of European highways for adopting connected, automated and electric vehicles; iv) connected roadway classification system; and v) road infrastructure ready for mixed vehicles traffic flows. Based on the methodologies, a questionnaire was prepared and applied to several specialists who work in the transportation area, whether in public management, consulting, academia, and private enterprise. A general and a specific analysis was performed according to the segments in which the specialists work. The most important components in the general vision of the specialists were: connectivity, security, physical infrastructure, digital infrastructure, vehicle automation, and user. Based on these components, the need arises to discuss what would be the path to be followed to implement and classify an Intelligent Road Corridor (IRC) in Brazil.

Concessions are a type of partnership with the private sector increasingly used by different governments to carry out road infrastructure works. These highly complex projects are characterized by a high concentration of investments in the first few years of cash flow, which requires raising a relevant portion of debt in the financial market. Understanding the importance of capital costs and setting up an adequate financial strategy is a critical success factor to be considered by public entities that act in the structuring of projects of this nature. In this context, an integrative literature review was carried out on the theme of financing road PPP projects, encompassing the main concepts and advantages of this modality, the main related parties and their interests in this process, and the different alternatives for public and private financing applicable. The elements found in the literature were compared with evidence and empirical data extracted from the Brazilian market of highway concessions. In the second part of the research, a questionnaire was applied to a group of specialists who work in banks that finance concession projects in Brazil. Based on the AHP methodology, a model was developed to measure the relative importance of the different aspects that make up the due diligence process of these financial institutions. 17 analysis criteria were compared, divided into 5 different dimensions. The results show that the most relevant factors considered by banks for granting loans are those related to project revenue generation estimates and the associated risk level, followed by the soundness of contractual clauses and the correct pricing of planned investments. These criteria account for more than 65% of the relevance of all aspects analyzed. The research results bring important indications about the main elements of the economicfinancial model that must be refined in order to enhance the financeability of road concession projects. In addition, through the integrative literature review, it was possible to list several possibilities for government participation in the financing of PPP projects, with the potential to be used in Brazil.

A locomoção por bicicleta está aumentando significativamente em todo o mundo e como um modo de transporte ativo o ciclismo pode potencialmente reduzir o congestionamento do tráfego e a poluição do ar. Além disso, promover um estilo de vida ativo pode melhorar a saúde pública e tornar as cidades mais amigáveis para as pessoas. No entanto, a quantidade de ocorrências e fatalidades envolvendo ciclistas ainda é preocupante. As Medidas de Segurança Substitutas (SSM) são indicadores promissores para avaliar a segurança no tráfego com base em conflitos, em vez de sinistros e acidentes. O termo "substitutas" é usado porque as medidas se baseiam em conflitos de tráfego e não em sinistros. Além disso, a triagem de rede garante uma identificação eficiente de locais perigosos para reduzir o número e a gravidade dos sinistros. Portanto, essa metodologia pode ser realizada tanto por uma abordagem reativa quanto por uma abordagem proativa. No que diz respeito às abordagens proativas, bicicletas instrumentadas com sensores tornaram-se cada vez mais úteis para pesquisas no campo da mobilidade. Por meio de um dispositivo de detecção portátil e multifuncional, é possível coletar dados de trajetória de bicicletas e a Distância Lateral de Ultrapassagem (LPD) usando variados sensores conectados a um sistema de banco de dados. Dessa forma, o presente estudo tem como objetivo estimar áreas de risco para a mobilidade ativa, aplicando algoritmos de aprendizado de máquina não supervisionado (K-Means e DBSCAN) com base em dispositivos sensores para a coleta de dados. Os resultados da Distância Lateral de Ultrapassagem (LPD) coletados entre bicicletas e veículos foram então relacionados aos dados do ciclista. Além da pesquisa de clusterização, foi realizada a correlação entre as características para identificar como os dados interagiam entre si. Alguns desses dados incluem velocidade, elevação do percurso, altitude, informações do acelerômetro e giroscópio a partir de uma coleta de dados naturalística na rua. Em relação aos dados gerais, 25% das leituras relativas à distância lateral de passagem têm menos de 136,36 cm. Quando o modelo de agrupamento é aplicado, 25% dessas leituras tiverem menos de 100,13 cm; para o segundo quartil, 50% tiveram menos de 193,69 cm. Isso indica uma LPD crítica para um dos grupos com 75 leituras, quando considerado o limite de 150 cm para a distância mínima de afastamento lateral estabelecida pela legislação brasileira. A metodologia foi aplicada a um estudo de caso em uma avenida do centro da cidade de Brasília, em torno do Parque da Cidade. Por fim, este estudo objetiva uma aplicação metodológica e analítica de segurança cicloviária orientada a dados através da utilização de algoritmos de aprendizado de máquina.

A condução sob a influência de álcool responde por quase um terço dos sinistros de trânsito ao redor do mudo. Embora os efeitos do álcool no corpo humano sejam bem conhecidos, os fatores que motivam as pessoas à prática desse comportamento ainda exigem uma análise mais detalhada. Através da Teoria do Comportamento Planejado de Ajzen – TPB (1991), buscou-se identificar as crenças salientes e avaliar os construtos que subjazem a essa conduta de dirigir sob efeito de álcool. Inicialmente, seguindo o método da TPB, um estudo piloto com questionário de perguntas abertas sobre crenças foi administrado a 192 motoristas no Distrito Federal. Posteriormente, no estudo principal, com perguntas fechadas para avaliar as crenças e os construtos básicos da TCP, 745 indivíduos responderam ao questionário. Um terceiro e último instrumento, tendo como objetivo o comportamento de fato, foi elaborado e aplicado, com 52 retornos. Os resultados da primeira etapa passaram por uma análise de conteúdo que identificou e codificou as crenças salientes comportamentais, normativas e de controle percebido que influenciam na decisão de beber e dirigir. No estudo principal, os dados passaram por procedimentos estatísticos por meio de correlações, análise fatorial exploratória e regressões. Os resultados apontaram que motoristas do sexo masculino, na medida que avançam em idade, são mais propensos a dirigir sob a influência de álcool. Além disso, verificou-se correlações significativas entre crenças comportamentais e de controle percebido face aos construtos da atitude (ATT) e do controle comportamental percebido (CCP) e, desses, na intenção (INT) de dirigir sob a influência de álcool. Normas subjetivas, embora importantes, tiveram menor impacto. Assim, sugere-se que possíveis medidas de intervenções pela segurança viária devem visar, principalmente, os construtos da ATT e do CCP, abordando medidas voltadas à percepção do risco, às consequências legais, aos efeitos do álcool no organismo, à influência de pessoas significativas, como vítimas e autoridades de trânsito, e ao fomento ao uso de meios alternativos de locomoção quando sob efeito de álcool.

The objective of the thesis was to estimate the relationships established by residential mobility for understanding travel behavior. To this end, a conceptual model and a measurement instrument on residential mobility were developed to encompass the individual decision to stay in their current residence or change residence. The choice of residential location was analyzed in relation to the built environment and residential self-selection to understand the choice of transportation mode. In addition to the conceptual contribution, the methodological procedures of the research instrument involved analysis, translation, adaptation, and scientific validations to carry out an empirical application. The model was tested using structural equation modeling with 1,418 individuals in the Federal District, Brazil, through two distinct samples: those who stayed and those who changed residence in the last 5 years. The results obtained from both samples indicate that the characteristics of the built environment, such as urban structure, distance, and neighborhood, have a greater influence on the choice of residential location than behavioral factors related to residential self-selection, such as travel abilities, preferences, and needs. The results presented in both samples suggest that the built environment does not influence the choice of transportation mode, and residential self-selection does not influence the choice of residential location. Furthermore, for the sample of individuals who stayed in their residence, it was found that residential self-selection does not influence the decision to stay. In conclusion, for both samples, the decision to stay or change residence and the choice of residential location influence the choice of transportation mode.

People seldom travel only for moving around, they travel to attend their needs. In this sense single trips models, largely used on transportation planning, cannot fully represent reality. Models that use chains of trips, also known as tours, may help to solve this problem and better understand mode choice, since literature shows that tours that start by car usually do not use other modes. There is an interest in determining what variables are significant for modeling trip chaining and therefore mode choice. One common set of variables used in travel behavior models is the built environment ones, which effects are understood for mode choice. However, despite the efforts researchers still have to come to an agreement of the effects of land use on trip chaining. This study aims to further understand the influence of the built environment on two aspects of travel behavior: trip chaining complexity and mode choice across different populations. Using the data from the Federal District Urban Mobility Survey, six groups were formed, three income classes divided between workers and non-workers resulting in a total of 30,871 tours. Due to the different nature of data, the trip chaining behavior were modeled through a ordered logit model and mode choice modeled through a multinomial logit model. Results showed that the built environment had an effect on trip chaining, but that different classes are affected by different variables. Difference on the average income of the destination of a person and his home is significant across the studied populations, indicating that the trip chaining process might be sensible to how much one can afford in a region. The mode choice results showed that tours with more stops had a greater probability of using the car over transit, but it had no significant effects on choosing an active mode over car. By analyzing both results the research concludes that large trip chains may not be only in the realm of motorized travel.

Urban Freight Transport (UFT) is responsible for moving goods in urban areas to meet the needs of citizens, therefore, UFT is essential for economic development. However, the UFT contributes to the generation of adverse impacts on society and the environment, such as congestion and pollution. Since the UFT is, simultaneously, one of those responsible for maintaining the urban lifestyle and generating negative externalities that impact the lives of residents, this research assesses how the urban infrastructure and the negative externalities generated by the UFT influence the quality of life of the residents. It is assumed that urban infrastructure is the responsibility of public managers and influenced by the business community; negative externalities arise from UFT operations and are influenced by government actions. Finally, infrastructure and externalities influence the quality of life of residents. The research is carried out in Brasília, Brazil, a planned city with problems of heavy traffic, noise, air pollution and congestion. A questionnaire with 22 questions was developed to assess whether the negative externalities generated by the UFT, and urban infrastructure influence the perception of the population's quality of life. The questionnaire provides data to validate seven hypotheses developed for this research, presented via the web to the residents of the Plano Piloto of Brasília. 403 valid responses were obtained, which had their data analyzed by Structural Equation Modeling (SEM), with the Partial Least Squares approach (PLS-SEM). The Training Composite model is configured with two independent variables (Public Management and Private Management), two intervening variables (Infrastructure and Externalities) and one dependent (Quality of Life). The results show that the population's quality of life is more influenced by the negative externalities generated by the UFT than by the urban infrastructure, the first being negative and the second positive. In addition, the results show a strong influence of public and private management on externalities and, on the other hand, low influence of infrastructure. Finally, a roadmap indicates the issues that should be prioritized to improve the quality of life of the Plano Piloto de Brasília, which point out as a priority the capacity of the roads, the proper use, and the number of spaces for loading and unloading.