Dissertation Defence: Developing and Deploying an Activity-Based Model to Test Telecommuting and Active Transportation Policies, and Temporal Transferability

Bijoy Saha, supervised by Dr. Mahmudur Fatmi, will defend their dissertation titled “Developing and Deploying an Activity-Based Model to Test Telecommuting and Active Transportation Policies, and Temporal Transferability” in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering.

An abstract for Bijoy Saha’s dissertation is included below.

Examinations are open to all members of the campus community as well as the general public. Please email mahmudur.fatmi@ubc.ca to receive the Zoom link for this exam.

Abstract

Activity-based models (ABMs) are widely used for transportation policy analysis. However, they often lack sensitivity in capturing behavioural responses to temporally sensitive policies such as telecommuting strategies and active transportation plans such as bike user-targeted interventions. The temporal transferability of ABMs for generating practical future scenarios to assess policies is also challenging, as travel patterns are changing due to behavioural adaptation to virtual mobility. With this motivation, this dissertation adds capacity to the existing ABM literature by accommodating behavioural heterogeneity across population groups and improving policy testing exercises. Specifically, a nested modelling structure is developed to investigate how time use, land use and travel decisions interact with each other and vary across telecommuters, commuters, and non-workers. The findings reveal significant differences across these population groups; for example, telecommuters are flexible in changing their travel choices under varying circumstances, whereas commuters are mostly inflexible. This dissertation also investigates bike users’ behaviour, specifically their destination choice, in two stages: choice-set generation and model estimation. Bike users experience distinct spatio-temporal and physical strength constraints than other mode users, which are incorporated in choice-set generation using time-space prism concept. Destination choice is estimated using a logit model, and the findings highlight that bike users may travel farther for simple tours with one stop, but prefer shorter distances for each trip leg in complex tours with multiple stops. The behavioural models are integrated into an activity-based travel demand and network simulation framework, Activity SIMulator (ASIM), thus introducing policy-testing capabilities. The dissertation deploys ASIM to assess the impact of telecommuting, revealing that a higher share of telecommuters increases non-work trips to destinations near urban centers during off-peak hours, resulting in peak spreading. In addition, bike user-targeted scenarios examining car and bike ownership show that regulating both can significantly reduce car share and increase bike network utilization. The dissertation implements a temporal transferability methodology by introducing sensitivity to changes in activity participation and their frequency. Forecasting future traffic under different behavioural assumptions demonstrates that neglecting behavioural changes may lead to biased estimates. Thus, this dissertation supports evidence-based policymaking with methodological innovations, behavioural insights, and policy-oriented scenario analysis.