Traffic management models that include route choice form the basis of traffic management systems. High-fidelity models that are based on rapidly evolving contextual conditions can have significant impact on smart and energy efficient transportation. Existing traffic/route choice models are generic and are calibrated on static contextual conditions. These models do not consider dynamic contextual conditions such as the location, failure of certain portions of the road network, the social network structure of population inhabiting the region, route choices made by other drivers, extreme conditions, etc. As a result, the model’s predictions are made at an aggregate level and for a fixed set of contextual factors. There is a clear need to develop traffic models that take into account local contexts and are closer to ground reality to provide government agencies the ability to make well-informed model-based decisions/policies. In this project: (1) used Immersive Virtual Environment (IVE) tools for generating context-aware and high-fidelity data related to drivers’ route choice behavior, (2) developed a novel approach for developing high-fidelity route choice models with increased predictive power by augmenting existing aggregate level baseline models with information on drivers' responses to contextual factors obtained from stated choice experiments carried out in an IVE through the use of knowledge distillation. To this end, the study used a virtual driving environment designed based on I-10 in Baton Rouge, LA. Five alternate routes were introduced to the participant. Ten experimental scenarios were conducted to produce initial data about drivers’ dynamic route choice behavior, given emerging contextual factors. Experimental results have demonstrated that the predictions of the augmented models produced by our approach are much closer to reality than that of the baseline. Our study demonstrates that existing route choice models based on econometric theories cannot accurately predict behavior in real world scenarios. For high-fidelity route choice models, one needs to combine existing route choice models with information about contextual factors gleaned from SCEs.
Mukhopadhyay, S., Zhu, Y., & Gudishala, R. (2019). Combining Virtual Reality and Machine Learning for Enhancing the Resiliency of Transportation Infrastructure in Extreme Events. Retrieved from https://repository.lsu.edu/transet_pubs/43