COVID-19 imposed travel restrictions have induced significant changes to our travel behaviour and daily life, such as work-from-home and online learning. The long-lasting nature of this pandemic might trigger a longer-term change in our behaviour after the pandemic, such as continued preference for work-from-home and e-learning. Such changes in work and learning arrangements do not only indicate a reduction in travel during the peak hour, it might also indicate a shift in travel to other times of day as well as changes in trip purposes and travel distances. For example, a worker telecommuting might spend the whole/part of the day at home and then go out to meet friends/family for dinner and do groceries from a store near the home while returning. Traditional four-stage travel demand models typically take the origin-destination (OD) matrix for the peak hours as input at an aggregate-level of temporal, spatial, and population resolution and do not necessarily accommodate the trip chaining behaviour of an individual. As a result, the behavioural changes associated with time-sensitive policies such as work-from-home and e-learning are not accommodated and/or reflected by these models. This demands the development agent-based transport simulation models adopting activity-based modelling technique. This study adopts an agent-based transport network simulation technique to generate 24-hour traffic for alternative work-from-home and online learning strategies. The model is calibrated and validated for the Central Okanagan region of British Columbia, Canada. Specifically, the open source Multi-Agent Transport Simulation (MATSim) model has been adopted, which was written using the Java programming language. The 24-hour travel schedule is developed adopting an activity-based modelling technique. The findings of this study will assist the governments and transit agencies in understanding the dynamics of travel behaviour and the consequent change in traffic patterns over the 24-hour for alternative work arrangement scenarios.
Keywords: Work-from-home, E-learning, Agent-based transport simulation model, Activity-based model, MATSim
