Omer Tatari
University of Central Florida
Civil Engineering

Final Report

View PDF|

Final Report Summary

View PDF|




Agent-Based Simulation for Investigating the Safety Concerns of Electric Vehicles in the US

An important concern of the transportation sector is that it depends heavily on fossil fuels and responsible for large amounts of environmental emissions. Therefore, many alternative fuel types have been tested for all on-road vehicle types. Pure electric powertrain vehicles are one of these alternatives and draw significant amount of attention from users, vehicle manufacturers, and government organizations since it could reshape the transportation industry with zero on-site emissions, specific infrastructure (charging station) requirements, and eliminating government's share from fuels. The market penetration of electric vehicles (EVs) is expected to increase rapidly for future years with fossil fuel price increase projections (Noori and Tatari 2016; Noori et al. 2015; Onat et al. 2014). Although National Highway Traffic Safety Administration (NHTSA) has not yet rated some of the proposed or already released EV models, some rated EV models present significantly high safety rates (i.e. Tesla Model S- overall 5-star, Chevy Volt-overall 5-star, Nissan-Leaf, overall 4-star) (NHTSA 2016). These safety test rates also indicate that EVs adopt advanced safety technologies. There is no additional safety regulation for EVs since they use similar body, weight, seating, etc. specifications with internal combustion engine (ICE) vehicles. However, there are safety concerns on EVs that silent engine pose treat for pedestrian and cyclist and even more for blind people. NHTSA's study concluded that hybrid-electric vehicle involve more crashes than ICE vehicle with pedestrians and cyclists (Hanna 2009). The most prevalent safety concern declared by the experts and researchers is that the EVs may suffer from frequent battery depletion, which may cause them to stop suddenly in the middle of the traffic. As opposed to safety concerns on EVs, some statistical observation studies in literature indicate that hybrid or EV drivers intend to be more careful or in other words less risk takers in traffic (Horswill and Coster 2002). Hence, it is critical to study the interactions and behaviors of EVs, drivers, pedestrians, and other agents simultaneously. This research proposes an agent-based simulation approach to create a virtual environment to model the interactions between different agents. Agent-based simulation approach is an ideal methodology to approach such a problem of safety where many agents, including ICE vehicle driver, EV driver, pedestrian, cyclist, blind pedestrian, government regulations, and vehicle manufacturers, simultaneously interact with each other. Historically collected crash data and driver characteristics with vehicle choices will be utilized for the foundation of this simulation. The developed simulation model will allow us to project crash rates, with the consideration of increasing number of EVs on the roads.

Supporting links:
TRID Record