Mode Choice Analysis Research Articles

Moving Leisure Travellers to Low-Carbon Transportation: A Mode-Choice Analysis in Western Australia

Moving Leisure Travellers to Low-Carbon Transportation: A Mode-Choice Analysis in Western Australia

Analysis of Access and Egress Mode Choice to High-Speed Railway Stations

Analysis of Access and Egress Mode Choice to High-Speed Railway Stations

Deep Neural Network for On-Demand Mobility Mode Choice Analysis: Issues on Adaptability and Interpretability

Deep Neural Network for On-Demand Mobility Mode Choice Analysis: Issues on Adaptability and Interpretability

Comparative mode choice analysis of university staff commuting travel preferences

Surveying citizen preferences on transportation modes when commuting is a major issue in urban transport planning. Most of the current methods approach the problem through the attributes of choices thus forecasting the demand indirectly. This paper aims to analyze a survey of commuting students and university staff by two direct preference models: the Analytic Hierarchy Process and the Best-Worst Method. Both techniques are based on pairwise comparisons; consequently, the commuting transport alternatives can be directly compared with each other, and the results are comparable, too. However, the two methods differ in the number and the nature of comparisons and in the consistency check, thus they can be regarded as competitors. A real-world case study on commuting student groups provides a better understanding of the proposed methodology. As a result, it can be stated that despite their low utilization in the transportation field, both the Analytic Hierarchy Process and the Best-Worst Method are applicable to mode choice preference surveys, and they produce comprehensive final outcomes. Therefore, the well-known tools of mode choice can be extended by Multi-Criteria-Decision-Making techniques to increase the efficiency of transport demand prediction. The extension is beneficial to avoid the bias of other methods in converting attribute evaluations to real mode choice decision, as both models, especially the Best-Worst approach, requires less cost and time than the mainstream techniques.

Passively wait for gridlock, or proactively invest in service? Strategies to promote car-to-transit switches among aspirational urbanites in rapidly developing contexts

As planners and policymakers in cities around the world struggle to attract and retain public transit users, this paper explores strategies to promote transit adoption in contexts where the odds are stacked against transit. Using travel behavior data from India's National Capital Region – one of the most congested metropolitan areas globally that is experiencing unprecedented growth in personal vehicle ownership and plummeting transit ridership – this study analyzes the choice of intra-urban (metro) rail over personal car for commute trips within a specific population sub-group that is fast adopting the personal car and exiting the transit market. The objective of this mode choice analysis that employs both logistic regression modeling and the propensity-score matching technique is to identify multi-modal service quality/performance factors that determine metro rail vs. car choice, and thereby recommend efficient and effective interventions for inducing car-to-transit switches in addition to retaining existing transit riders. Results suggest that increasing metro rail's travel time competitiveness relative to car (particularly for long distance commuters), service frequency and safety level, and raising car parking cost at the destination can increase the likelihood of choice of metro rail over car in the study context. Interestingly, increase in traffic congestion and travel time unreliability of the car mode are not expected to automatically boost the demand for metro rail use, all else equal. As the momentum towards more car adoption continues, this study shows that deteriorating traffic conditions may not push drivers out of cars; rather, public transit has to step up and pull drivers out of cars. For planners and policymakers, this study indicates that rather than passively waiting for driving conditions to worsen, they should proactively invest in transit service quality improvements. This study also highlights the importance of communicating service changes to the traveling public in order to efficiently translate interventions to behavior changes, given the dissonance between travelers' perceptions of multi-modal travel conditions and actual travel conditions. This paper further demonstrates that analysis of travelers' perception errors, including variation in error across travelers, is important for accurately modeling travel behavior changes in response to interventions. The findings add to the literature on mode choice analysis, and provide strategic advice for transit agencies in India and across other comparable contexts globally.

Mode Choice Analysis of Sustainable Public Transport in Bogor City of Indonesia

Public transport service scenarios must be well designed for public transport performance to be better and more sustainable. The aim of this study was to determine the percentage of the travelers move to public transportation in the Bogor City. The Stated Preference technique is used to know travelers responses to public transport service scenarios in utility functions. Furthermore, from the binomial logit model analysis obtained the probability of choice of public transportation mode. The results of the analysis have shown that the travelers responded positively to move to public transportation modes. In Corridor III, private car users and online-based motorcycle transport users choose the public transportation of non-economic medium bus type (scenario-2) with the greatest probabilities of 35.59 % and 61.02 %. In Corridor II, private car users choose the public transportation of non-economic big bus type (scenario-3) with the greatest probability of 36.31 %. Motorcycle users and online-based motorcycle transport users choose the public transportation of non-economic small bus type (scenario-1) with the greatest probabilities of 66.47 % and 70.64 %. The motorcycle users in Corridor II choose the public transportation in all scenarios with significant probabilities ranging from 62.03% to 66.47%. Keywords: mode choice analysis, public transportation, sustainable DOI: 10.7176/JESD/12-18-05 Publication date: September 30 th 2021

Multinomial Logit Utility Model for Tanta City Transportation System considering Ridesharing Mode

Recently, ridesharing has been noticed as an effective and sustainable mode of transport. In which, each passenger carries out a journey that benefits travelers and society greatly, such as reducing travel costs, reducing journey times, relieving road traffic congestions, preserving fuel and reducing air pollution. While the importance and efficiency of ridesharing, ridesharing among travelers has not been commonly utilized. This paper introduces an upgraded disaggregated utility model for a mode choice step in the four-step traffic demand model for transport of passengers in Tanta city - Gharbiya governorate. For the first time, the analysis in this paper includes ridesharing mode in addition to other three common modes; private vehicle, bus, and taxi . The aim of this research is to involve the ridesharing mode and new parameters in the mode choice analysis of the case study. This is done by including additional parameters such as travel cost and comfortable in addition to travel time. Based on surveys, the developed Multinomial logit utility models have been assessed using the McFadden pseudo R2 values. The values demonstrated high compatibility of results with real data as pseudo R2 values ranges between 0.2 and 0.4. In addition, all utility models developed for modes are found to have P-values less than 0.05 indicating the significance of the considered utility characteristics.

Autonomous vehicle adoption: use phase environmental implications

This paper investigates the environmental trade offs resulting from the adoption of autonomous vehicles (AVs) as a function of modal shifts and use phase. An empirical approach is taken to formulate a mode choice model informed from a stated preference (SP) survey conducted in Madison, Wisconsin. A life cycle analysis based on well-to-wheel model is then conducted to quantify the use phase environmental impacts across different categories. The mode choice analysis reveals the potential users of AVs and its attractiveness as a mode of transportation, ultimately competing with traditional modes available. This translates into modal shifts that are shown to result in an expected increase in environmental impacts across all studied categories: energy consumption (5.93%), greenhouse gas emissions (5.72%), particulate matter (6.80%), sulfur (6.85%) and nitrogen oxides (5.70%). The adoption of electric AVs (E-AVs) is then analyzed as an offsetting strategy to combat the increase in environmental impacts. The analysis reveals that E-AVs are capable of offsetting the foreseen impacts, yet their effectiveness is dependent on the electricity generation mix and adoption rate.

Analysis of travel mode choice and trip chain pattern relationships based on multi-day GPS data: A case study in Shanghai, China

As the propensity to link multiple intermediate stops in a trip chain (a sequence of journeys that starts and ends at home, includes visiting one or more locations) is more prevalent, the relationship between travel mode choice and trip chain pattern aroused the attention of academics. This paper examines two distinct structures to identify the decision process of travelers between travel mode choice and trip chain pattern: one structure in which trip chain pattern organization precedes travel mode choice, another structure in which travel mode choice decision precedes the organization of trip chain pattern. To accommodate multi-day behavioral variability and unobserved heterogeneity in personal characteristics ignored by traditional travel surveys, multi-day GPS data collected in Shanghai is employed to estimate these two structures within Nested Logit (NL) model. The Monte Carlo (MC) method simulates the switch of trip-chaining and mode choice under possible Transportation Demand Management (TDM) strategies based on estimation results. The findings of this study are as follows: (1) trip chain pattern decision precedes travel mode choice, which means trip-chaining is organized first and affects travel mode choice; (2) complex trip chain is related to higher automobile dependency, and it is a barrier to the tendency to adopt public transit; (3) people who generally travel by automobiles might switch to public transit when private cars are unavailable, and an increase in household bicycle ownership enhances competition between the bicycle and public transit which leads people to turn to cycling. These findings help implement TDM strategies to develop sustainable transportation systems and optimize the urban trip structure.

Examining the Effects of the Built Environment on Travel Mode Choice across Different Age Groups in Seoul using a Random Forest Method

It is important to analyze factors that influence travel mode choice and to predict individual mode choice because this shapes people’s movement and determines their level of mobility. While there have been studies investigating how built-environment elements are associated with travel mode choice, most efforts have neglected evaluating the heterogeneity of effects that the built environment has on travel mode choice across different age groups. This study aims to examine the effects of the built environment in influencing travel mode choice across age groups in Seoul, South Korea, using a random forest approach. Our random forest model demonstrates what factors are important and how they are associated with the effects on travel mode choice. As a result, the built environment has a greater impact on the subway selection for older adults than other age groups and the random forest approach captures non-linear relationships between certain predictors and travel mode choices. Applying this approach to the travel mode choice analysis, we can examine the heterogeneous effects of the built environment on travel mode choice across different age groups.

Decision tree analysis of commuter mode choice in Baguio City, Philippines

Transportation is a multidisciplinary system. Solving its issues would require the knowledge of social, economic, engineering, environmental, and technological disciplines. Emerging techniques used in problem-solving involve the use of machine learning techniques. In this study, a machine learning technique, decision tree, is used for mode choice analysis in Baguio City, Philippines. Using data from a household survey, the developed model uncovers the most significant factors affecting mode choice of residents in the city. The results highlight the role of income, which is related to the individuals’ career level and stage in life. Interestingly, a mid-level income group seems to be highly inclined towards private vehicle use. To conclude, the authors note that the primary advantage of a decision tree is its simplicity and straightforward results interpretation, which is paramount in policymaking. For future work, the authors recommend exploring larger decision tree models for mode choice and conducting a validation interview of the insights obtained from the study.

Mode choice analysis among motorcycle and Trans Koetaradja urban bus and its contributing factors using revealed preference (RP) data

The random utility maximization (RUM) is a widely used principle to study customer behaviors derived from econometric theory. The mode choice problem in transportation planning has been intensively studied to explore the choice behavior of travelers. In the demand analysis, market shares among transport modes are substantially used for public transport planning and operation. Therefore, this study is a preliminary analysis of the modeling complex mode choice behaviors among motorcycle and bus system so-called Trans Koetaradja by considering specific attributes of mode, households, and individual sociodemographic attributes. The developed model of mode choice is used to explore the effect of specific attributes of mode, households, and individual sociodemographic attributes on traveler’s mode choice behaviors and its contributing factors. Using the Revealed Preference (RP) data collected in Banda Aceh, we formulate Binary Logit (BL) to identify and evaluate the effects specific attributes of mode, households, and individual sociodemographic on mode choice. These results confirm that there is a substantial contribution of the specific mode attributes of the individual in determining selected mode choice. The specific attribute of motorcycle mode preference such as travel time (TT) and distance of destination (DD). Findings further confirm that individual sociodemographic variable including dummy young Age (DYA), dummy has elementary education background (DEB), dummy has low-medium income (DIN), dummy owned motorcycle (DOM) and dummy has driver license (DDL) have a statistically significant alert to the selecting mode choice of motorcycle.

Using neural nets to predict transportation mode choice: Amsterdam network change analysis

In the Amsterdam metropolitan area, the opening of a new metro line along the north–south axis of the city has introduced a significant change in the region’s public transportation network. Mode choice analysis can help in assessment of changes in traveler behavior that occurred after the opening of the new metro line. As it is known that artificial neural nets excel at complex classification problems, this paper aims to investigate an approach where the traveler’s transportation mode is predicted through a neural net, trained on choice sets and user specific attributes inferred from the data. The method shows promising results. It is shown that such models perform better when it is asked to predict the choice of mode for trips which take place on the same underlying transportation network as the data with which the model is trained. This difference in performance is observed to be especially high for trips from and to certain areas that were impacted by the introduction of the north–south line, indicating possible changes in behavioural patterns, entailing interesting possible directions for further research.

Applying transfer learning and various ANN architectures to predict transportation mode choice in Amsterdam

For long, statistical models have been used for transportation mode choice analysis, due to their ability to extract economic information from the model parameters. Recently, the application of Artificial Neural Nets to predict transportation mode choice is gaining ground, partly due to efforts that have led to an improved interpretability of this class of models. In this development, various innovations have been suggested concerning Neural Net architecture and hyperparameter tuning. Building on this, this paper investigates 3 similar Neural Net architectures to be applied to data from an Amsterdam case study. This data has been collected in 3 waves. Between the first and second collection period, the public transportation network in Amsterdam changed. A transfer learning approach is suggested to improve models that were trained on a single wave of data. Based on the test loss of the models from the transfer learning experiments, we conclude that this is a promising technique to use in this context, since it has shown to improve model performance.

Mode Choice Analysis of Trips of Students and Staff in Education Hub Using Multinomial Logit Model

Mode Choice Analysis of Trips of Students and Staff in Education Hub Using Multinomial Logit Model

Modeling Social Distance and Activity-Travel Decision Similarity to Identify Influential Agents in Social Networks and Geographic Space and Its Application to Travel Mode Choice Analysis

Because humans are social beings, people are members of social networks and interact with other members. As a result of social interaction, people can be influenced by the behavior of others. The present study addresses conformity behavior in activity-travel decisions, implying that in making such decisions people mimic the behavior of other members of their social networks. The presence of conformity behavior in social networks implies that sustainable behavior can be dispersed through networks. Therefore, knowing which people in a network are influential can help make a sustainable transportation policy more effective. In particular, information about the topology of social networks and geographical distribution can help maximize the policy’s spill-over effects in social and geographic spaces. This study suggests a framework to locate influential agents in relation to activity-travel decisions using three procedures: (1) estimating social distance associated with similarity in activity-travel decisions, (2) identifying influential agents by measuring centralities, and (3) exploring the spatial and activity-travel characteristics of the influential agents. The suggested framework is applied using the travel mode choices of people who had recently taken trips by road beside/alongside the Han River in Seoul, South Korea.

A Nested Ensemble Approach with ANNs to Investigate the Effect of Socioeconomic Attributes on Active Commuting of University Students.

Analysis of travel mode choice is vital in policymaking and transportation planning to comprehend and forecast travel demands. Universities resemble major trip attraction hubs, with many students and faculty members living on campus or nearby. This study aims to investigate the effects of socioeconomic characteristics on the travel mode choice of university students. A nested ensemble approach with artificial neural networks (ANNs) was used to model the mode choice behavior. It was found that students generally prefer motorized modes (bus and car). A more detailed analysis revealed that teenage students (aged 17–19 years) had an approximately equal probability of selecting motorized and non-motorized modes. Graduate students revealed a higher tendency to select motorized modes compared with other students. The findings of this study demonstrate the need to promote non-motorized modes of transport among students, which is possible by providing favorable infrastructure for these modes.

Estimating bicycle demand in an aggressive environment

The perception of the environment where we move can be a key factor in choosing a transport mode for our daily trips. For example, many cities lack adequate infrastructure or experience environmental conditions that do not favor cycling. We examine the potential demand for using the bicycle as a transport alternative in trips to college in Barranquilla, Colombia, a city with fairly aggressive environmental conditions for cycling. We designed a Stated Choice experiment geared to students and staff living within three kilometers of two college campuses in the city and traveling by bus or bike. We also considered people’s perceptions in an effort to understand the reasons why only 2% of Barranquilla’s daily trips are made by bicycle. Our mode choice analysis incorporated two latent variables (unsafety/insecurity and convenience), in addition to typical level-of-service attributes, such as travel time and cost. We found that the perception of unsafety/insecurity has an important negative effect on the probability of using bicycle; the possibility of suffering an accident or being mugged, weighs much more than the possible economic savings and improvements in health associated with using the bicycle. In terms of convenience, our results suggest that it could be just as important to have showers on campus as bike lanes available on the route. Also, investment in facilities such as free parking for bikes, and strategies such as planting trees in bike paths and sidewalks to provide shade, could significantly encourage cycling.

Prediction and behavioral analysis of travel mode choice: A comparison of machine learning and logit models

Some recent studies have shown that machine learning can achieve higher predictive accuracy than logit models. However, existing studies rarely examine behavioral outputs (e.g., marginal effects and elasticities) that can be derived from machine-learning models and compare the results with those obtained from logit models. In other words, there has not been a comprehensive comparison between logit models and machine learning that covers both prediction and behavioral analysis, two equally important subjects in travel-behavior study. This paper addresses this gap by examining the key differences in model development, evaluation, and behavioral interpretation between logit and machine-learning models for mode-choice modeling. We empirically evaluate the two approaches using stated-preference survey data. Consistent with the literature, this paper finds that the best-performing machine-learning model, random forest, has significantly higher predictive accuracy than multinomial logit and mixed logit models. The random forest model and the two logit models largely agree on several aspects of the behavioral outputs, including variable importance and the direction of association between independent variables and mode choice. However, we find that the random forest model produces behaviorally unreasonable arc elasticities and marginal effects when these behavioral outputs are computed from a standard approach. After the introduction of some modifications that overcome the limitations of tree-based models, the results are improved to some extent. There appears to be a tradeoff between predictive accuracy and behavioral soundness when choosing between machine learning and logit models in mode-choice modeling.

Analysis of travel characteristics and access mode choice of elderly urban rail riders in Denver, Colorado

With the proportion of the elderly increasing worldwide, significant research efforts have been devoted to understanding the travel characteristics of the ageing population. Our research contributes to a limited number of studies that have investigated how the elderly (65+) access urban rail in comparison to younger adults (19–64). Using descriptive and econometric analyses of on-board survey data from Denver, Colorado, we respond to the following two research questions: (1) How do trip characteristics, including access and egress mode, trip purpose, choice of station, and travel time of elderly urban rail riders differ from those of their younger counterparts? (2) Do the determinants of access mode choice, such as built environment and individual characteristics, impact differently the choices of elderly and younger adult riders? Our results indicate that the elderly urban rail users differ from the average elderly population in the US in terms of their employment status (46% of the elderly Denver rail riders are employed). Consequently, dissimilarities in travel behavior between elderly and younger adult riders are on average lower than expected, although important differences are still observed. Results from a multinomial probit model on access mode choice for trips between home and a rail station show that younger adults and employed elderly travelers place higher importance on travel time when choosing a mode to access a station compared to non-employed elderly travelers. On the other hand, the built environment and vehicle ownership have a significantly higher influence on access mode choice for elderly compared to younger adult travelers.