This research presents the first North American route-choice model for cyclists developed from a large sample of GPS data. These findings should encourage all interested municipalities to implement cycling as part of their transportation planning by determining key designing and planning factors to encourage cycling. The analysis is based on processing revealed preference data obtained from 415 self-selected cyclists in Waterloo, Ontario, which corresponded to 2000 routes. Cyclists' route decisions were modeled using multinomial logit framework of discrete choice theory. The main finding involved in capturing two different behaviour groups, namely experienced and inexperienced cyclists. This was subsequently reflected in the two developed models. The key factors impacting route-choice were found to be trip length, speed, volume, bicycle lane presence and percent of uphill gradient that cyclists face. The predictive power of the best model was 65%. The outlier analysis found that the relative significance of uphill gradient coefficient in one circumstances and perhaps the exclusion of unobserved variables, in other circumstances could be the cause why probability of actual choice was not predicted by both models all the time. In addition, this research involved in the development of a transferability study involving route-choice modeling for cyclists. The analysis is based on the revealed preference data obtained from 255 self-selected cyclists in Peel Region, Ontario, which corresponded to 425 unique routes. The choice set contained actual routes and a combination of alternatives obtained by labeling and impedance rules. The transferability of Waterloo's model to Peel Region was 37%. This means that cyclists behaviour in the Peel Region can be predicted correctly by travel length, bicycle lane presence and percent of uphill gradient for every third cyclist.

Drivers need to make judgements of physical relationships related to driving speed, such as mean speed, risks, travel time and fuel consumption, in order to make optimal choices of vehicle speed. This is also the case for the general public, politicians and other stakeholders who are engaged in traffic issues. This thesis investigates how drivers’ judgements of travel time (Study I and II), fuel consumption (Study III) and mean speed (Study IV) relate to actual physical measures. A cognitive time-saving bias has been found in judgements of travel time. The time saving bias implies that people overestimate the time saved when increasing speed from a high speed and underestimate the time saved when increasing speed from a low speed. Previous studies have mainly investigated the bias from a cognitive perspective in questionnaires. In Study I the bias was shown to be present when participants were engaged in a driving simulator task where participants primarily rely on perceptual cues. Study II showed that intuitive time saving judgements can be debiased by presenting drivers with an alternative speedometer that indicate the inverted speed in minutes per kilometre. In Study III, judgements of fuel consumption at increasing and decreasing speeds were examined, and the results showed systematic deviations from correct measures. In particular, professional truck drivers underestimated the fuel saving effect of a decrease in speed. Study IV showed that subjective mean speed judgements differed from objective mean speeds and could predict route choice better than objective mean speeds. The results indicate that biases in these judgements are robust and that they predict behaviour. The thesis concludes that judgements of mean speeds, time savings and fuel consumption systematically deviate from physical measures. The results have implications for predicting travel behaviour and the design of driver feedback systems. / Förare bör göra bedömningar som relaterar till hastighet, såsom bedömningar av medelhastighet, risk, restid och bränsleåtgång. Dessa bedömningar är nödvändiga för att föraren ska kunna välja en optimal hastighet, men också för att allmänheten, politiker och andra intressenter som är involverade i trafikfrågor ska kunna fatta välgrundade beslut. Denna avhandling består av fyra delstudier där förares bedömningar av restid (Studie I och II), bränsleåtgång (Studie III) och medelhastighet (Studie IV) studeras i relation till faktiska fysikaliska mått. Tidigare enkätstudier har påvisat ett kognitivt bias i tidsvinstbedömningar vid höga och låga hastigheter som påverkar mänskligt beteende. Studie I visade att detta bias också förekommer i en primärt perceptuell motorisk uppgift där förarna i studien kör i en körsimulator. Studie II visade att dessa intuitiva tidsbedömningar kan förbättras genom att köra med en alternativ hastighetsmätare i bilen som indikerar den inverterade hastigheten i minuter per kilometer istället för hastigheten i kilometer per timme. I Studie III undersöktes bedömningar av bränsleåtgång vid hastighetsökningar och hastighetssänkningar, och resultaten visar att bedömningarna systematiskt avviker från faktisk bränsleåtgång. Ett intressant resultat var att lastbilsförare i allmänhet underskattade bränslebesparingen som kan göras till följd av en hastighetssänkning. Studie IV visade att subjektiva bedömningar av medelhastighet som avviker från objektiva medelhastigheter kan predicera vägval, vilket tyder på att systematiska fel i dessa bedömningar är robusta och kan predicera vägval. Sammanfattningsvis visar avhandlingen hur bedömningar av medelhastighet, tidsvinst och bränsleåtgång systematiskt avviker från fysikaliska mått. Resultaten har betydelse för modellering av resebeteende och design av förarstödssystem. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Submitted. Paper 3: Submitted.</p>

Thesis (Ph. D.)--Universität Bremen, 2007. / Includes bibliographical references (p. 155-168).

Thesis (MA)--Stellenbosch University, 2001. / ENGLISH ABSTRACT: A shortest or most economical route can easily be determined by using a geographical information system (GIS). Unfortunately, most systems compute distances in two dimensional space. As computer-technology moves towards three dimensional applications, it is essential that GIS keeps up with this trend. In this research, the network module of Arc View (using Avenue) is customized, so that topographical slope is considered in determining the shortest or most economical route. Two buttons were added to the normal Arc View interface. By doing it this way, the user has the full functionality of Arc View, as well as the use of the new application. One button initiates a dialogue for capturing the economic parameters (fuel efficiency, oil usage, tyre usage, maintenance costs and capital costs) of a vehicle. The other button selects a route network and uses a vehicle's economic parameters (as determined by the user) to calculate a most economical route. This thesis describes the procedure, logic and methodology followed in adding a most economical route-selection function to Arc View. It also demonstrates the importance of incorporating three dimensional space for determining a most economical route. The new function currently calculates a most economical route, based on vehicle running costs for Heavy Goods Vehicles (HGV's). The application performs satisfactorily, but there is scope for further development and refinement, both of the economical formulae for computing costs as well as of the graphic user interface (GUl). The flexibility of the system can be enhanced by providing for additional classes of vehicles. / AFRIKAANSE OPSOMMING: 'n Kortste of mees ekonomiese ritroete kan maklik met behulp van 'n GIS (Geografiese Inligtingstelsel) vasgestel word, maar die meeste stelsels bereken afstande in 'n plat vlak (in twee-dimensionele ruimte). Soos die rekenaartegnologie ontwikkel, word meer drie-dimensionele ruimtelike toepassings geskep, dus moet GIS-tegnologie ook toenemend die derde dimensie inkorporeer. In hierdie navorsing is Arc View se netwerk module met Avenue aangepas dat dit topografiese helling outomaties inreken by die bepaling van 'n kortste of mees ekonomiese roete. Twee knoppies is tot die normale Arc View koppelvlak bygevoeg. Deur dit so te doen, het die gebruiker toegang tot die volle funksionaliteit van Arc View en dié van die nuwe funksie. Een knoppie inisieer die koppelvlak waarmee die ekonomiese parameters (brandstof verbruik, olie verbruik, band verbruik, kapitaal koste en onderhoudskoste) van 'n voertuig opgestel word. Die ander knoppie selekteer 'n padnetwerk en gebruik 'n voertuig se ekonomiese parameters (soos gedefinieer deur die gebruiker) om 'n mees ekonomiese roete vas te stel. Hierdie tesis beskryf die prosedures, logika en metodologie waarvolgens die nuwe roeteseleksie funksie by Arc View geïnkorporeer is. Dit het ook gedemonstreer dat dit noodsaaklik is om drie-dimensionele ruimte by die bepaling van 'n mees ekonomiese roete in te sluit. Die nuwe funksie bepaal tans 'n ekonomiese roete gebaseer op die voertuig-loopkoste van swaarvoertuie. Dit funksioneer bevredigend, maar daar is steeds moontlikhede vir verdere ontwikkeling en verfyning, beide van die ekonomiese kosteberekeningsformules en die gebruikers-koppelvlak. Deur ook vir ander klasse voertuie voorsiening te maak kan die plooibaarheid van die stelselook uitgebrei word.

Um melhor uso da infraestrutura da rede de transporte é um ponto fundamental para atenuar os efeitos dos congestionamentos no trânsito. Este trabalho utiliza aprendizado por reforço multiagente (MARL) para melhorar o uso da infraestrutura e, consequentemente, mitigar tais congestionamentos. A partir disso, diversos desafios surgem. Primeiro, a maioria da literatura assume que os motoristas aprendem (semáforos não possuem nenhum tipo de aprendizado) ou os semáforos aprendem (motoristas não alteram seus comportamentos). Em segundo lugar, independentemente do tipo de classe de agentes e do tipo de aprendizado, as ações são altamente acopladas, tornando a tarefa de aprendizado mais difícil. Terceiro, quando duas classes de agentes co-aprendem, as tarefas de aprendizado de cada agente são de natureza diferente (do ponto de vista do aprendizado por reforço multiagente). Finalmente, é utilizada uma modelagem microscópica, que modela os agentes com um alto nível de detalhes, o que não é trivial, pois cada agente tem seu próprio ritmo de aprendizado. Portanto, este trabalho não propõe somente a abordagem de co-aprendizado em agentes que atuam em ambiente compartilhado, mas também argumenta que essa tarefa precisa ser formulada de forma assíncrona. Além disso, os agentes motoristas podem atualizar os valores das ações disponíveis ao receber informações de outros motoristas. Os resultados mostram que a abordagem proposta, baseada no coaprendizado, supera outras políticas em termos de tempo médio de viagem. Além disso, quando o co-aprendizado é utilizado, as filas de veículos parados nos semáforos são menores. / A better use of transport network infrastructure is a key point in mitigating the effects of traffic congestion. This work uses multiagent reinforcement learning (MARL) to improve the use of infrastructure and, consequently, to reduce such congestion. From this, several challenges arise. First, most literature assumes that drivers learn (traffic lights do not have any type of learning) or the traffic lights learn (drivers do not change their behaviors). Second, regardless of the type of agent class and the type of learning, the actions are highly coupled, making the learning task more difficult. Third, when two classes of agents co-learn, the learning tasks of each agent are of a different nature (from the point of view of multiagent reinforcement learning). Finally, a microscopic modeling is used, which models the agents with a high level of detail, which is not trivial, since each agent has its own learning pace. Therefore, this work does not only propose the co-learnig approach in agents that act in a shared environment, but also argues that this taks needs to be formulated asynchronously. In addtion, driver agents can update the value of the available actions by receiving information from other drivers. The results show that the proposed approach, based on co-learning, outperforms other policies regarding average travel time. Also, when co-learning is use, queues of stopped vehicles at traffic lights are lower.

O estímulo aos modos ativos de viagem através de modificações na forma urbana - infraestrutura, edificações e atividades - é objeto do planejamento das cidades contemporâneas desde que os problemas do transporte motorizado se revelaram críticos e contrários à ideia de sustentabilidade urbana. Congestionamentos, poluição, custos monetários e sociais elevados estão associados ao modelo de urbanização e mobilidade da maioria das cidades, especialmente nos países em desenvolvimento, onde a explosão da frota motorizada é concomitante à aceleração dos processos de urbanização e espalhamento urbano. Mais recentemente, a revelação de relações de causalidade entre o excesso de utilização dos veículos particulares e problemas de saúde crônica das populações urbanas levou a um crescimento das pesquisas e iniciativas de estímulo às caminhadas como modo de transporte. Também o interesse pela qualidade da experiência da vida na cidade veio somar-se a este corpo de conhecimento, trazendo as ideias de vitalidade urbana e urbanidade para junto dos estudos de caminhabilidade. O interesse extrapolou a análise de demanda agregada que visa o incremento do modo a pé e passou a dar atenção aos caminhos pelos quais as pessoas se movimentam, conectando origens e destinos nos interior das cidades. Este trabalho procura avançar neste aspecto ao propor um método de avaliação dos atributos da forma urbana baseado no monitoramento de caminhantes com dispositivos de posicionamento global (GPS) e modelos de escolha discreta. Um estudo na cidade de Porto Alegre acompanhou indivíduos durante os anos de 2011 a 2014 e, através da representação de diversos atributos urbanos em ambiente SIG, associou as trajetórias realizadas com as características da forma da cidade, concluindo que elas influenciam a utilidade percebida das alternativas de caminho e, portanto, atuam sobre o processo decisório dos pedestres. Para além da simples distância ou declividade, outras características como o tamanho dos trechos, a hierarquia das vias, a presença de prédios marcantes e espaços abertos e a densidade de edificações ao longo dos eixos revelaram-se influentes neste processo. Poucas intersecções e cruzamentos, predomínio de vias amigáveis ao pedestre e edificações arranjadas de forma menos densa são alguns dos atributos que apresentaram maior relevância para a decisão de “por onde ir” no âmbito deste estudo. A amostra relativamente pequena e a concentração das viagens no entorno do Parque Farroupilha e do campus da UFRGS não permitem generalizar os resultados. Porém, o estudo pode ser considerado válido enquanto exploração, pois constrói uma metodologia que pode ser ampliada e aplicada em outros contextos. Além disso, os resultados revelam particularidades da realidade local que parecem indicar a existência de diferenças comportamentais significativas em relação às cidades do primeiro mundo, tornando-o promissor como instrumento de suporte a políticas e projetos de mobilidade urbana sustentável no Brasil. / Since motorized transport problems have proved to be critical and contrary to the concept of urban sustainability, the idea of increasing the active travel modes through changes in urban form is a key subject of today´s mobility agenda. Traffic congestion, air pollution and severe monetary and social costs are associated with the current patterns of urbanization and mobility, especially in developing countries, where an explosive motorized fleet growth occurs simultaneously to an acceleration of urbanization and sprawl processes. Most recently evidences of a causal relationship between massive use of private vehicles and chronic health disorders have led to an increase in research about walking as an effective and clean mode of transportation. Also, the interest about the quality of life experience in the city came to add up to this body of knowledge, bringing in ideas of livability to walkability studies. Beyond the aggregate demand studies that aim to increase the walking mode share, there is now a growing interest on more localized aspects of the walking phenomenon - the routes - trying to understand the ways in which people travel on foot when connecting origins and destinations. This work proposes a method based on assessing data from actually taken walking trips using GPS devices and on modeling pedestrian´s choice behavior using discrete choice models. A study in the city of Porto Alegre, south of Brazil, followed 82 individuals for three consecutive days and, through the representation of several layers of urban data in a GIS environment, associated their trajectories with the main urban form attributes to allow the modeling experiment. The results show that the built environment features play an important role as a decision attribute, producing perceived utility/disutility on the decision-makers´ minds. They indicated that, in addition to the basic travel effort attributes such as trip distance or street slope, other factors such as the straightness of the trip, the road hierarchy, the presence of busy intersections, landmark buildings, noticeable public spaces and the density of buildings along the walking stretches indeed influence the route choice. The relatively small sample size and the spatial clustering of trips around the city´s central area doesn´t allow to the generalization of results. However, the study can be taken as a valid exploratory analysis, since it builds up a methodology that can be expanded and applied in other urban contexts. Furthermore, the results reveal some particular local features that indicate the existence of significant behavioral differences from the developed cities where previous similar studies were performed. These qualities make the proposed framework a promising decision support tool for sustainable urban mobility projects in Brazil.

The design of new facilities - buildings, shopping centers, public transport stations, airports, or intersections of urban roads - should consider delays resulting from intense pedestrians’ flows in order to make its' operation more efficient. The general objective of this doctoral thesis is to propose a simulation model to represent pedestrians’ behavior in urban environments. Simulation models should allow planning these environments in order to provide greater levels of comfort and safety for the pedestrian. Agent-based abstraction has been widely used for pedestrian modeling, mainly due to its capacity to represent complex entities. Agent-based models represent agents’ decision-making ability based on their profile and perception over the environment. One of the most important pedestrians’ activities is the route choice. This document describes the development of a route choice model based on friction forces. The route cost calculation considers a balance between distance and the impedance generated by other pedestrians. Simulations runs shown that pedestrians choosing longer routes can have similar or better travel times. The ability of choosing not only the shorter route brings more realistic behaviors for the pedestrians’ representation, especially with small differences in route lengths and higher congestion. On the proposed model agents were modeled with partial knowledge of the network conditions. The knowledge was limited considering the pedestrian estimated field of view. In the real world it is not possible to know the network state before turning the corner. The model was validated and calibrated with real data. Calibrating a pedestrian route choice model is a complex task mainly for two reasons: (i) Many factors interfere on pedestrians’ route choice; (ii) data collection is difficult. To overcome these difficulties real pedestrians were studied in a controlled environment. An experiment was set up inside the university campus. After the calibration process the model was able to simulate a real scenario. Proposed model was applied to simulate a shopping mall environment. Simulate the pedestrians shopping behavior is particularly complex once route choice in shopping malls may be defined by a number of causal factors. Shoppers may follow a pre-defined schedule; they may be influenced by other people walking, or may want to get a glimpse of a familiar shopping. Analysis from simulations indicates that the agents’ behavior provides a promising approach for real case applications.

Um melhor uso da infraestrutura da rede de transporte é um ponto fundamental para atenuar os efeitos dos congestionamentos no trânsito. Este trabalho utiliza aprendizado por reforço multiagente (MARL) para melhorar o uso da infraestrutura e, consequentemente, mitigar tais congestionamentos. A partir disso, diversos desafios surgem. Primeiro, a maioria da literatura assume que os motoristas aprendem (semáforos não possuem nenhum tipo de aprendizado) ou os semáforos aprendem (motoristas não alteram seus comportamentos). Em segundo lugar, independentemente do tipo de classe de agentes e do tipo de aprendizado, as ações são altamente acopladas, tornando a tarefa de aprendizado mais difícil. Terceiro, quando duas classes de agentes co-aprendem, as tarefas de aprendizado de cada agente são de natureza diferente (do ponto de vista do aprendizado por reforço multiagente). Finalmente, é utilizada uma modelagem microscópica, que modela os agentes com um alto nível de detalhes, o que não é trivial, pois cada agente tem seu próprio ritmo de aprendizado. Portanto, este trabalho não propõe somente a abordagem de co-aprendizado em agentes que atuam em ambiente compartilhado, mas também argumenta que essa tarefa precisa ser formulada de forma assíncrona. Além disso, os agentes motoristas podem atualizar os valores das ações disponíveis ao receber informações de outros motoristas. Os resultados mostram que a abordagem proposta, baseada no coaprendizado, supera outras políticas em termos de tempo médio de viagem. Além disso, quando o co-aprendizado é utilizado, as filas de veículos parados nos semáforos são menores. / A better use of transport network infrastructure is a key point in mitigating the effects of traffic congestion. This work uses multiagent reinforcement learning (MARL) to improve the use of infrastructure and, consequently, to reduce such congestion. From this, several challenges arise. First, most literature assumes that drivers learn (traffic lights do not have any type of learning) or the traffic lights learn (drivers do not change their behaviors). Second, regardless of the type of agent class and the type of learning, the actions are highly coupled, making the learning task more difficult. Third, when two classes of agents co-learn, the learning tasks of each agent are of a different nature (from the point of view of multiagent reinforcement learning). Finally, a microscopic modeling is used, which models the agents with a high level of detail, which is not trivial, since each agent has its own learning pace. Therefore, this work does not only propose the co-learnig approach in agents that act in a shared environment, but also argues that this taks needs to be formulated asynchronously. In addtion, driver agents can update the value of the available actions by receiving information from other drivers. The results show that the proposed approach, based on co-learning, outperforms other policies regarding average travel time. Also, when co-learning is use, queues of stopped vehicles at traffic lights are lower.

The design of new facilities - buildings, shopping centers, public transport stations, airports, or intersections of urban roads - should consider delays resulting from intense pedestrians’ flows in order to make its' operation more efficient. The general objective of this doctoral thesis is to propose a simulation model to represent pedestrians’ behavior in urban environments. Simulation models should allow planning these environments in order to provide greater levels of comfort and safety for the pedestrian. Agent-based abstraction has been widely used for pedestrian modeling, mainly due to its capacity to represent complex entities. Agent-based models represent agents’ decision-making ability based on their profile and perception over the environment. One of the most important pedestrians’ activities is the route choice. This document describes the development of a route choice model based on friction forces. The route cost calculation considers a balance between distance and the impedance generated by other pedestrians. Simulations runs shown that pedestrians choosing longer routes can have similar or better travel times. The ability of choosing not only the shorter route brings more realistic behaviors for the pedestrians’ representation, especially with small differences in route lengths and higher congestion. On the proposed model agents were modeled with partial knowledge of the network conditions. The knowledge was limited considering the pedestrian estimated field of view. In the real world it is not possible to know the network state before turning the corner. The model was validated and calibrated with real data. Calibrating a pedestrian route choice model is a complex task mainly for two reasons: (i) Many factors interfere on pedestrians’ route choice; (ii) data collection is difficult. To overcome these difficulties real pedestrians were studied in a controlled environment. An experiment was set up inside the university campus. After the calibration process the model was able to simulate a real scenario. Proposed model was applied to simulate a shopping mall environment. Simulate the pedestrians shopping behavior is particularly complex once route choice in shopping malls may be defined by a number of causal factors. Shoppers may follow a pre-defined schedule; they may be influenced by other people walking, or may want to get a glimpse of a familiar shopping. Analysis from simulations indicates that the agents’ behavior provides a promising approach for real case applications.

Esta dissertação busca identificar aspectos carentes de melhorias na modelagem de pedestres. A modelagem do comportamento de pedestres se inicia pelo entendimento de seu processo decisório, entendendo como ele se desloca, realiza escolha de rotas, interage com outros pedestres, veículos, obstáculos, etc. Os modelos de simulação de pedestres estão hoje aptos a representar de forma bastante realista o deslocamento dos mesmos, entretanto, a simulação conjunta com veículos apresenta ainda alguns aspectos fracos devido as grandes diferenças de abordagem na modelagem dos dois modos. A representação da travessia de pedestres nos modelos de simulação tradicionais apresenta limitações que podem impactar nos resultados gerados pelos simuladores. Este trabalho enumera diversos comportamentos de pedestres decorrentes da interação com veículos, no momento da travessia, que geralmente não estão presentes nos simuladores. Uma vez identificados estes comportamentos, uma pesquisa foi realizada com usuários do sistema viário, com idades entre 22 e 60 anos, para avaliar o quanto estes comportamentos são frequentes e importantes na estimativa de tempos de viagem dos pedestres. O comportamento indicado pelos entrevistados como mais impactante nos tempos de viagem foi o de “busca por brecha em caminhada”, onde o pedestre percorre trechos na lateral da via, enquanto observa possíveis brechas na corrente de tráfego para realizar sua travessia. O referido comportamento foi então modelado e agregado a um modelo de simulação de pedestres. Os resultados mostraram que a inclusão do novo comportamento provoca redução significativa dos tempos médios de viagem dos pedestres e que a simulação pode ser mais condizente com o comportamento real de pedestres em diversos ambientes urbanos. Para representar o comportamento de um pedestre em um ambiente urbano, é preciso também estudar como ele escolhe suas rotas. No processo de escolha de rotas, o pedestre é influenciado por diversos fatores, como hábitos pessoais, o número de cruzamentos, níveis de poluição e de ruído, segurança, abrigo de condições climáticas ruins, e estimulação do ambiente. Para representar o comportamento de escolha de rota dos pedestres, foi desenvolvido um modelo que considera a interação entre pedestres como uma impedância alterando a rota do pedestre. O estudo foi inspirado por equações de forças de atrito, considerando que pedestres tendem a evitar passar próximo de outros pedestres com elevada velocidade relativa. Para escolher uma rota o pedestre realiza uma ponderação entre a impedância e a distância a ser percorrida. O modelo foi capaz de reproduzir comportamentos emergentes da interação entre os agentes, permitindo concluir que as equações de forças de atrito adotadas nesta modelagem podem ser uma abordagem válida na representação da escolha de rotas de pedestres, podendo também ser uma forma indireta de avaliação de atrasos. / This dissertation aims to identify aspects in need of improvement in modeling pedestrians. The modeling of pedestrian behavior begins by understanding their decision making process, understanding how people move, make route choice, interact with other pedestrians, vehicles, obstacles, etc. Simulation models of pedestrians are able to represent the way they move quite realistically, however, the combined simulation of pedestrians and vehicles still presenting some poor aspects due to the wide differences in the modeling approach of the two modes. The pedestrian road crossing representation in the traditional simulation models has limitations that may impact on the results generated by the simulators. This dissertation lists several behaviors arising from the interaction of pedestrians with vehicles at road crossing situations, which are generally not present in the simulators. Having identified these behaviors, a survey was conducted with pedestrians, aged between 22 and 60 years old, to evaluate how these behaviors are frequent and important to estimate travel times. The behavior indicated by the interviewees as more impactful in the travel times was the “search for a gap while walking”, describing the pedestrian that walks laterally to the road, in the sidewalk, at the same time that is trying to cross the road, looking for gaps in the traffic stream. Such behavior was modeled and then aggregated at a simulation model of pedestrians. Results showed that the inclusion of the new behavior causes significant reduction in average travel time for pedestrians and that the simulation can be more consistent with the actual behavior of pedestrians in different urban environments. To represent a pedestrian behavior in an urban environment, it is also necessary to study how he chooses their route. At the route choice process, the pedestrian is influenced by several factors, such as personal habits, the number of road crossings, levels of pollution and noise, safety, shelter from bad weather, and other stimulation of the environment. In order to represent the pedestrians route choice behavior was devised a model that considers the interaction between pedestrians as an impedance to alter pedestrians route. The study is inspired by friction forces equations, considering that pedestrians avoid passing near other pedestrians with high relative velocity. To choose a route a pedestrian consider a balance between the impedance and the path length. The model is able to reproduces emergent behavior between agents, allowing the assumption that friction equations adopted in this modeling may provide a suitable approach to route choice behavior and can also be used as an indirect measure of pedestrians delay.