A spatial agent-based model (ABM) simulation of individuals’ walking behavior in a city demonstrated that ABMs can effectively formalize the dynamic processes behind and facilitate understanding of individuals’ walking behavior.
ABMs are computational models used to simulate individuals’ actions and their interactions with each other and their environments. The authors assigned characteristics to each individual in their simulation, including age, walking ability, attitude toward walking and home location, and observed the individuals traveling for work, leisure and basic needs. Feedback mechanisms updated individuals’ attitudes toward walking on a daily basis. The authors also used the ABM to explore how distributions of non-household locations and safety affect socioeconomic differences in walking.
According to this simulation, whether and how much people walk varies with the distance they need to travel, their walking ability and their attitude about walking. Individuals’ attitudes about walking change over time depending on the previous day’s walking distance and experiences, walking with others and others’ attitudes toward walking.
The model used in this study could be improved with the inclusion of gender differences, variation in the aesthetics of a city, more complex walking trips and the effects of long-term habit-formation. With modifications, ABMs could be used to study walking behavior and identify interventions to increase walking.