How On-demand Food Delivery impacts traffic network during pandemic? 

Abstract
The market for on-demand food delivery has increased considerably throughout the world, especially during the pandemic. It is crucial for transportation and environmental agencies to understand how on-demand food delivery reshape the travel patterns of people, and impact the Vehicle Miles Traveled (VMT), fuel use and emissions in the transportation system. However, the lack of public data on the operation of on-demand food delivery companies makes it challenging to quantify the impact of on-demand delivery in the real world transportation network. In this research, we formulate a Pick-up and Delivery Problem with Time Window (PDPTW) model and modified the adaptive large neighborhood search (ALNS) algorithm to solve the large-scale on-demand delivery problem at city-level. We then use a daily activity generation tool, CEMDAP, to create a simulation scenario to model the on-demand food delivery behaviors based on real world network, restaurant locations and population demographics in the City of Riverside, CA. The system-level evaluation shows that on-demand food delivery has great potential to reduce dining-related VMT while reducing fuels and emissions in the transportation network. The simulation results also indicate that the total dining-related VMT during COVID-19 decreased by 24.92% compared to the pre-COVID period, and the corresponding environmental impact also reduced.