Pedestrian Route Choice Behavior Research Articles

Exploring the Pedestrian Route Choice Behaviors by Machine Learning Models

To investigate pedestrian route choice mechanisms from a perspective distinct from that employed in discrete choice models (DCMs), this study utilizes machine learning models and employs SHapley Additive exPlanations (SHAP) for model interpretation. The data used in this paper come from several pedestrian flow experiments with two routes, which were recorded by UAV. Our findings indicate that logistic regression (similar to a binary logit model) exhibits good computational efficiency but falls short in predictive accuracy when compared to other machine learning models. Among the 12 machine learning models assessed, by calculating the new indicator named OP, we find that eXtreme Gradient Boosting (XGB) and Light Gradient Boosting (LGB) strike the best balance between accuracy and computational efficiency. Regarding feature contribution, our analysis reveals that bottlenecks exert the most significant influence on pedestrian route choice behavior, followed by the time it takes pedestrians to return from the end of the route to the origin (reflecting pedestrian characteristics and attitudes). While the pedestrian density of the shorter route contributes less compared to bottlenecks and return time, it exhibits a threshold effect, meaning that once the density of the shorter route surpasses a certain threshold, most pedestrians opt for the longer route.

What do pedestrians consider when choosing a route? The role of safety, security, and attractiveness perceptions and the built environment during day and night walking

Increasing the use of non-motorized modes of transport, such as walking, is a worldwide objective aimed at improving the sustainability of cities. However, pedestrians may not choose to walk if the infrastructure fails to meet their needs or if they hold unfavourable perceptions regarding the built environment (BE). The current study aims to identify the significance of route attributes and perceptions of attractiveness, safety, and security at the route level, which influence pedestrians' preferences for last-mile route choices. A cross-sectional design was employed, utilizing a questionnaire comprising a stated preference (SP) experiment and a perception survey. The study considered theory-informed attributes influencing pedestrian route choice preferences, including: (1) individual-level determinants, (2) physical-level determinants, and (3) time of day. Two separate models were developed, considering the time of day, to examine the differences in trade-offs within pedestrians' route choice preferences between day and night. The results revealed that both the BE and perceptions of the BE play a crucial role in determining pedestrian route choice behaviour. Pedestrians showed a preference for routes fully encompassed by mixed or residential land uses during the daytime. The presence of vacant land along the walking route significantly decreased the likelihood of choosing a route at night. Generally, pedestrians favoured shorter walking times, lower posted speed limits, and comfortable walkway grades in their routes. Female pedestrians tended to avoid routes that were not well-lit and pleasant at night. Lowering roadway speed limits emerged as a strategy to encourage walking in suburban areas. The findings of this study hold the potential to play an essential role in the development of effective policy initiatives targeted at pedestrians in cities.

Understanding pedestrian route choice behavior in the continuous space: Diversity and equilibrium

Exploring general rules from diverse motion choices in the real world is a pivotal area of study that deepens our understanding of pedestrian behavior. In this research, we conduct pedestrian circle antipode experiments, from which we extract pedestrians' trajectories from repeated trials. We focus on the symmetric features identified from these experiments and employ the K-Medoids trajectory clustering method alongside several behavioral indicators, namely route length, travel time, and comfort level, to distinguish and compare different trajectory route choice patterns. In doing so, we delve into the heterogeneous behaviors exhibited by pedestrians within a single experiment and the stochastic behaviors seen across different experiments for a single pedestrian. Additionally, we establish a linear model to scrutinize the equilibrium relationship among these behavioral indicators, and further deploy the Gaussian process to shed light on the equilibrium characteristics inherent in pedestrian route choice behavior. Our findings suggest that despite the inherent heterogeneity and stochastic nature of individual behaviors, pedestrian route choice overall operates in a dynamic equilibrium state. These insights offer valuable implications for the design of facilities in transport hubs and crowd management strategies in social activities, potentially leading to more effective and organized public spaces.

Capturing positive network attributes during the estimation of recursive logit models: A prism-based approach

Although the recursive logit (RL) model has been recently popular and has led to many applications and extensions, an important numerical issue with respect to the computation of value functions remains unsolved. This issue is particularly significant for model estimation, during which the parameters are updated every iteration and may violate the feasibility condition of the value function. To solve this numerical issue of the value function in the model estimation, this study performs an extensive analysis of a prism-constrained RL (Prism-RL) model proposed by Oyama and Hato (2019), which has a path set constrained by the prism defined based upon a state-extended network representation. The numerical experiments have shown two important properties of the Prism-RL model for parameter estimation. First, the prism-based approach enables estimation regardless of the initial and true parameter values, even in cases where the original RL model cannot be estimated due to the numerical problem. We also successfully captured a positive effect of the presence of street green on pedestrian route choice in a real application. Second, the Prism-RL model achieved better fit and prediction performance than the RL model, by implicitly restricting paths with large detour or many loops. Defining the prism-based path set in a data-oriented manner, we demonstrated the possibility of the Prism-RL model describing more realistic route choice behavior. The capture of positive network attributes while retaining the diversity of path alternatives is important in many applications such as pedestrian route choice and sequential destination choice behavior, and thus the prism-based approach significantly extends the practical applicability of the RL model.

How do street attributes affect willingness-to-walk? City-wide pedestrian route choice analysis using big data from Boston and San Francisco

This study adds to the nascent but growing literature on the use of big data for pedestrian route choice analysis. We explore behavioral preferences for various route attributes in Boston, MA using a large dataset of GPS trajectories (n = 11,165) sourced from a third-party smartphone app. Although the data are anonymized and limit our exploration of user heterogeneity, the sample size and area coverage are both much larger than seen in most previous studies. We estimate route choice preferences using a path size logit model, and operationalize the coefficients for policy-making through ‘willingness-to-walk’ measures. The value of these measures is demonstrated through an application of computing pedestrian accessibility to transit stations. Additionally, we compare our findings to a previous study in San Francisco, CA using similar data and methods, and previous literature to explore similarities and differences in pedestrian route choice behavior across major metropolitan areas more generally. While our findings can inform walkability policy and practice on several counts, we recommend future efforts to focus on supplementing this study by surveying hard-to-reach populations for more equitable policy-making.

Preference for Street Environment Based on Route Choice Behavior While Walking.

This study aimed to better understand the relationship between the street environment and walking behavior by deciphering the pedestrians' street environment preference based on their route choice behavior while walking. The route data of 219 residents were collected using an unobtrusive tracking method and subjected to binary logistic regression models to analyze the pedestrian route choice behavior. The results revealed that except for the walking distance, the trip purpose and travel status are the potential factors influencing the route choice of pedestrians. Furthermore, it was revealed that on-street parking, garbage bins, and streetlights could influence the pedestrians to select longer distance routes. In addition, pedestrians were more likely to select the shortest distance route when they were engaged in leisure activities with an accompanist. The findings of this study would offer insights, from different perspectives, into the micro-scale street environment and the walking behavior of pedestrians.

Effect of the Street Environment on Walking Behavior: A Case Study Using the Route Choice Model in the Chunliu Community of Dalian.

To better comprehend the relationship between the environment and walking, this study developed a conceptual framework that explained the association between the street environment and the route choice behavior of pedestrians. We collected the route choice data of 219 residents of the Chunliu community in Dalian and used a conditional Logit model to analyze the factors influencing route choice behavior to explain how the street environment affected pedestrians' walking habits and induced them to choose longer or more complicated routes for their activities. We found that sidewalk and driveway width, garbage bins, green spaces, the characteristics of street walls, the proportion of facilities could influence pedestrians' walking habits and compel them to choose longer and more complex routes. This study would provide new insights into walking characteristics and offer policy recommendations to the government on improving the street environment.

Simulating the effects of redesigned street-scale built environments on access/egress pedestrian flows to metro stations

In urban renewal processes, metro line systems are widely used to accommodate the massive traffic needs and stimulate the redevelopment of the local area. The route choice of pedestrians, emanating from or going to the metro stations, is influenced by the street-scale built environment. Many renewal processes involve the improvement of the street-level built environment and thus influence pedestrian flows. To assess the effects of urban design on pedestrian flows, this article presents the results of a simulation model of pedestrian route choice behavior around Yingkoudao metro station in the city center of Tianjin, China. Simulated pedestrian flows based on 4 scenarios of changes in street-scale built environment characteristics are compared. Results indicate that the main streets are disproportionally more affected than smaller streets. The promotion of an intensified land use mix does not lead to a high increase in the number of pedestrians who choose the involved route when traveling from/to the metro station, assuming fixed destination choice.

Potential-based three-dimensional route choice model for pedestrian evacuation on terraced stands

This study proposes a potential-based three-dimensional cellular automata model to describe the route choice behavior of pedestrians while evacuating terraced stands. The model takes into account the pedestrians’ states of sitting, standing, and moving, and the possible transitions on terraced stands. The proposed potential field algorithm reflects the influence on route choice behavior of heterogeneous heights, route distance, pedestrian congestion, and route capacity. Three parameters related to the pedestrians’ behaviors in the spaces with heterogeneous heights are then analyzed to demonstrate the ability of the model to reproduce pedestrian route choice behaviors on terraced stands. Finally, empirical data observed in a terraced stand of a stadium are applied for further validation. The results show that the proposed model performs consistently well with the observed data.

Simulation of pedestrian route choice with local view: A potential field approach

We investigate the route choice behavior of pedestrians under the condition of local view by using a field observation in a subway station and a microscopic pedestrian model with discrete space representation. The local view of a pedestrian means that the pedestrian's field of vision can be blocked by obstacles. Several typical phenomena with respect to behavior affected by the local view of pedestrians are observed through recording the movement of pedestrians in the subway station. Based on these observed phenomena and behaviors, a route choice model considering the effect of local view of pedestrians is developed. In the model, an algorithm for dividing feasibly vision areas is formulated. Two algorithms for computing the potential fields are then developed based on the results of dividing feasibly vision areas. By numerical experiments, the simulation results of the model are compared with the empirical observations to validate the model.

Analysis of Built Environment Influence on Pedestrian route choice behavior in Dutch Design Week using GPS Data

Visitors not only have specific destinations targeting the Dutch Design Week (DDW) exhibitions distributed all over the city, but also visit the city in between exhibition activities. The mixed environment makes modeling behavior of DDW visitors more complex than shoppers and tourisms only. This research pays special attention to the influence of built environment on pedestrian route choice. The built environment includes building and transportation infrastructure. GPS tracking data and social demographic information were collected during the event. Multinomial logit model and path size logit model are used to analysis route choice behavior. The results show that some built environment factors have significant influence on route choice. Shops are more attractive for aged visitors. Females prefer shorter routes more. In big event, the alternative routes with more sharing links could increase the possibility to choose.

A robust potential-based route choice model for simulating pedestrian evacuation

In this paper, a cellular automata model for pedestrian evacuation is presented. Special attention is paid to a phenomenon wherein an artificial and allegedly unrealistic bias in route choice is introduced. This phenomenon is caused by the so-called gray areas that affect the calculation of the potential fields which dictate pedestrians’ movements between lattice sites. In this paper, the concept of gray areas is formulated and the influence of the gray areas on the route choice is investigated for the first time. In brief, gray areas are those specific zones that should not influence the route choice behavior of pedestrians because they in fact cannot improve or deteriorate the evacuation safety, comfort and efficiency at all. To make the cellular automata model more robust to the gray areas, a new model is proposed based on a previous one, in which three factors including route distance, pedestrian congestion and route capacity are considered in the computation of the potential field. By changing the definition of route capacity from an exit dependent route capacity to be a path dependent one, the proposed model is not only comparable to the previous model in the sense of generating a wide range of route choice modes, but also more robust than the previous one when applied in the scenario containing gray areas. Such robustness is reflected in two aspects. First, the route choice patterns generated by the proposed algorithm are free from artificial and allegedly unrealistic bias in scenario with gray areas. Second, the route choice patterns generated by the proposed algorithm are less influenced by the extended gray areas and are closer to the optimal route choice. These two aspects are further confirmed by a comparison study with a finely validated perception-based model. Moreover, sensitivity analysis reveals that the proposed model is able to generate a wider range of route choice patterns. Finally, the results from the proposed model and empirical data are compared to show the effectiveness of the model in reproducing the route choice patterns observed in the reality.

Comparing the route-choice behavior of pedestrians around obstacles in a virtual experiment and a field study

Pedestrians often need to decide between different routes they can use to reach their intended destinations, both during emergencies and in their daily lives. This route-choice behavior is important in determining traffic management, evacuation efficiency and building design. Here, we use field observations and a virtual experiment to study the route choice behavior of pedestrians around obstacles delimiting exit routes and examine the influence of three factors, namely the local distance to route starting points and the pedestrian density and walking speeds along routes. Crucially, both field study and virtual experiment consider the same scenario which allows us to directly assess the validity of testing pedestrian behavior in virtual environments. We find that in both data sets the proportion of people who choose a closer exit route increases as the difference in distance between exit route starting points increases. Pedestrians’ choices in our data also depend on pedestrian density along routes, with people preferring less used routes. Our results thus confirm previously established route choice mechanisms and we can predict over 74% of choices based on these factors. The qualitative agreement in results between the field study and the virtual experiment suggests that in simple route-choice scenarios, such as the one we investigate here, virtual experiments can be a valid experimental technique for studying pedestrian behavior. We therefore provide much-needed empirical support for the emerging paradigm of experiments in virtual environments.

Objective vs. subjective measures of street environments in pedestrian route choice behaviour: Discrepancy and correlates of non-concordance

Difficulties in obtaining detailed street environment data is identified as a major obstacle in pedestrian route choice studies. Scholars have used either objectively or subjectively measured street environment data without testing their suitability as a substitute for each other in the route choice literature. This study aims to overcome these gaps by investigating the relevance of subjective data as a less-expensive proxy for objective data—together with identifying the factors affecting the degree of disconcordance between them. Subjective street environment data was collected from 178 pedestrians in Brisbane, Australia. Participants were intercepted and requested to draw their walking routes. They were asked to rate the importance of different street environment attributes influencing the chosen route. A range of objective attributes of these routes was derived through virtual audit and spatial analyses. The concordance of 13 seemingly related street environment variables was tested using ‘Kappa coefficient’ and ‘% agreement’ methods. Additionally, 13 multinomial logistic regression models were estimated, one for each variable, to identify different factors affecting the level of disconcordance. Results indicate a relatively poor agreement between objective and subjective attributes. However, an objective provisioning of some attributes can improve the perception of pedestrians about these attributes. Estimated regression models show that all groups are to some extent at-risk of being mismatched, suggesting the need for population-based policy interventions to improve perceptions. Findings highlight that subjective measures capture different construct of the street environment than those measured objectively, hence, using these two measures are not comparable.

Analysis of Route Choice for Pedestrian Two-Stage Crossing at a Signalized Intersection

Studying pedestrians’ twice-crossing behavior is of great significance to enhance safety and efficiency for pedestrians at signalized intersections. However, limited attention has been paid to analyze and model pedestrians’ behavior patterns of twice crossing. The purpose of this paper is to determine pedestrians' route choices for twice crossing at a signalized intersection, focusing on the waiting position (to cross the street) and walking route. A goal-oriented and time-driven model was proposed to analyze pedestrians’ twice-crossing behavior at signalized intersections, where the two directions have different pedestrian signal timing. A video-recording method was used to collect field data in order to obtain pedestrian preferences in choosing a walking route. It was found that pedestrians in the two directions present different preferences toward walking route, in waiting position, directional change and route type. The results showed that the proposed model is effective in simulating pedestrian route-choice behavior of twice crossing. This research provides a theoretical basis for identifying pedestrian movement intention, optimizing signal timing, and improving pedestrian infrastructure at signalized intersections.

Shortest path distance vs. least directional change: Empirical testing of space syntax and geographic theories concerning pedestrian route choice behaviour

Existing knowledge on the impact of built environment (BE) on route choice behaviour is doubtful due to an unresolved tension between two schools of thought. One represented BE geographically and found that most people tend to choose the shortest route. The other represented BE topologically and showed that least directional change is a key determinant of route choice. How do pedestrians make a trade-off between these two factors in route choice? This question is answered using walking route data of 178 pedestrians in Brisbane, Australia. Their reported routes were examined against the corresponding shortest path and least directional change routes using the percent overlap method. The effects of 25 additional BE factors were also estimated in a conditional logit model. The results reveal that: (a) Together distance and direction are likely to explain 53% of route choice decisions; (b) Individually, distance and direction are likely to explain 34% and 46% of route choice decisions respectively; (c) 28% of the chosen routes satisfied both distance and direction criteria. These findings confirm the claims of both schools, but also point that the least directional change route is a preferred option, and that pedestrians tend to minimise both criteria if they can.

Structured random walk parameter for heterogeneity in trip distance on modeling pedestrian route choice behavior at downtown area

A nested logit model of mode and walk route choice behavior in a downtown area is developed. The model samples alternatives at the route choice level, which is a lower level of the nesting structure. Alternative routes are drawn using the random walk method. A structured random walk parameter based on trip distance is proposed in order to consider the heterogeneity in trip distance, and a structured scale parameter based on trip distance is also applied for the nested logit model. Dataset is obtained by using mobile phones with GPS functions to track the trajectories in a downtown area at Nagoya, Japan. The results suggest that the proposed structured random walk parameter provides more efficient parameter estimates. Also, heteroscedasticity in the error terms for trips of various distances is found to be significant. Empirical findings include that older pedestrians prefer main shopping streets, and that roadside restaurants affect the route choice differently on weekdays and holidays. The different effects of the roadside environments according to the pedestrian’s attributes and the day of the week may indicate a need for more appropriate policies for specific pedestrian categories and days of the week.

Pedestrian route choice by elementary school students: the role of street network configuration and pedestrian quality attributes in walking to school

This study examines the extent to which street network configuration is associated with path selection by pedestrians. The aim is to better understand how the spatial layout of the street network affects pedestrian route choice behavior, controlling for pedestrian quality attributes. Randomly selected 6th, 7th, and 8th grade students (ages 12–14) from 15 elementary schools in Istanbul, Turkey were asked to draw their routes walking between home and school. Preliminary findings suggest that when all areas are merged into a single set, global choice, which measures the degree to which a street segment is located in between the shortest paths connecting all pairs of origins and destinations, and metric reach, which measures the street length that is accessible within a walking range, are significantly associated with route selection jointly with the distance to school, sidewalk width, distribution of pedestrians as well as the number of non-residential land-uses. However, directional reach, which measures the extent to which streets are sinuous or aligned, appears to be a strong correlate of route selection in the analysis of individual areas. From a design policy point of view, designing better connected street networks with reduced directional distance between home and school might serve as supporting navigation choices and walking behavior.

A potential field approach to the modeling of route choice in pedestrian evacuation

This study proposes a potential field algorithm for formulating pedestrian route choice behavior during evacuation in individual-based models with discrete space representation. The potential field reflects the effect of the route distance, pedestrian congestion and route capacity on route choice. Numerical simulations show that the developed model can reproduce more route choice modes in a scenario compared with several existing models. Three groups of pedestrian evacuation experiments are conducted and the proposed model reproduces pedestrian route choices effectively.

Simulation and experimental results of evacuation of pedestrian flow in a classroom with two exits

A microscopic pedestrian model based on cellular automata is proposed and three groups of experiments on pedestrian evacuation from a double-exit room are conducted to investigate the route choice of pedestrians during evacuation. In the model, the route-choice behavior of pedestrians is affected by the route distance from the pedestrians to the exit, the capacity of the frontal route, and the repulsive force between pedestrians. Through the analysis of the video recordings, certain conclusions related to the route choice of pedestrians in these experiments are obtained. Model parameters are calibrated by the experimental data. Simulation results indicate that the model can reproduce the evacuation behavior of pedestrians in the room effectively and evacuation time is an increasing linear function of student number. The present study is helpful for devising evacuation strategies and schemes in buildings that are similar to the room.