This special issue of Transportation Letters: The International Journal of Transportation Research represents a selection of notable papers which were presented at The 13th International Conference on Travel Behavior Research held from July 15th through 20th in Toronto, Canada. The papers represent a cross-section of the state-of-the-art in using agent-based microsimulation techniques in transportation modeling. The papers each deal in some way with various aspects of agent-based modeling, focusing either on new ways of incorporating agent-behaviors in microsimulation models or new methods for modeling and representing agent-behaviors, or in some cases both.
Agent-based microsimulation modeling is a topic that has received much attention recently in transportation research. Agent-based modeling techniques have found use in such disparate fields as economics, logistics, epidemiology, sociology and many others, and many of the techniques developed for models in these various fields have found use in travel demand and travel behavior modeling. This growth in interest has occurred as new simulation techniques and theoretical underpinnings have advanced the state of the art in general activity-based modeling, enabled primarily by the exponential growth in computational power to perform such simulations in recent decades. Agent-based microsimulation modeling is a broad topic that weaves together developments from complex adaptive systems research, artificial intelligence, cognitive and decision sciences, and many others into a methodology for representing the complex behaviors and interactions of individual agents, and the emergent effects that these actions and interactions can have on systems in which the agents operate (Macal and North Citation2010). This focus on individual agent behaviors and the effects they create when interacting with other agents lends itself naturally to the travel demand modeling field where individual travelers interact with each other in a system (the transportation network) and those interactions lead to system level effects (e.g. congestion, pollution, etc.) of interest to the modeler. Such use of agent-based microsimulation techniques can be seen in many of the activity-based travel demand modeling frameworks developed in recent years. A primary example is MATSIM (Balmer et al Citation2006)– which is the basis of two of the papers in this issue.
The first paper in this issue is ‘An Agent-Based Cellular Automaton Cruising-For-Parking Simulation’ by Andreas Horni, Lara Montini, Rashid A. Waraich and Kay W. Axhausen. This paper documents the development and implementation of a parking microsimulation intended to be implemented in the MATSIM agent-based modeling framework. The framework presented in the paper implements various agent behaviors relating to parking, including parking type choice, searching behaviors and parking lot choice. These behaviors guide the individual agents in the parking search process as they compete with other agents for limited parking locations.
The next paper also relates to work on the MATSIM framework. ‘Including joint trips in a multi-agent transport simulation’ by Thibaut Dubernet and Kay W. Axhausen provides an illustration of an agent-based algorithm for generating joint trips for multiple agents. A fundamental notion used in the simulation is that of the "clique" amongst which joint plans are generated and optimized. The paper details how such joint plan generation and optimization can be included in the general MATSim utility maximization approach.
The third paper in the issue, ‘Agent-based Modeling of Cognitive Learning of Dynamic Activity-Travel Patterns’ by Sehnaz Cenani, Theo A. Arentze and Harry J. P. Timmermans, is a unique contribution to travel behavior research which incorporates many ideas from spatial cognition and learning into a key aspect of travel demand modeling, namely spatial decision making. The paper details the current state of research in spatial cognition and the behavioral processes associated with it, such as mental map formation and updating. The paper describes a framework for modeling the process of spatial learning and sets out a pathway for incorporating such a framework into larger agent-based travel demand models.
The final two papers, ‘An Investigation of Household Interactions in Daily In-Home and Out-of-home Maintenance Activity Participation and Social Behaviour in Cairo, Egypt’ and the related paper ‘Modelling Individuals’ Attitudes towards Joint out-of-home Activity Participation with Household Members- Mosa and El Sawey’ were both written by Ahmed I. Mosa and Mohamed El Esawey. Each paper focuses on the important aspect of interactions amongst household members, and are both derived from an activity diary survey collected in Cairo, Egypt which had as a focus the individual participation in and attitudes towards joint activities. The first paper utilizes mixed logit models to investigate household interactions in in-home and out-of-home maintenance activity participation and social behavior. The paper looks at key drivers of what causes individuals to choose to participate in maintenance activities out of the home versus in-home and also the types of individuals with whom out of home activities tend to be conducted. The second of these papers focuses more on individuals' attitudes towards joint activity participation. The paper examines the role of an individuals’ attributes in forming their joint out of home travel habits and desires. Ordinal probit models are used to relate the individuals’ attitudes about joint activities to their socio-demographic and associated built environment and transportation features.
All of the papers in this issue focus in some way on understanding or implementing key agent behaviors which are currently or will soon be found in agent-based microsimulation models of travel demand. Much progress has been made in incorporating agent-based modeling techniques in the travel demand modeling field, however much still remains to be done in this area. It is my hope that this special issue will help further this process and serve as a useful resource.
Joshua A. Auld
Transportation Research and Analysis Computing Center, Argonne National Laboratory, Lemont, Illinois, USA
References
- Macal CM, North M.J. (2010). Tutorial on agent-based modelling and simulation. Journal of Simulation, 4, 151–162.
- Balmer M, Nagel K., Raney B. (2006). Agent-based demand modeling framework for large scale micro-simulations, Transportation Research Record, 1985, 125–134.