AI for Driving Simulator Vehicles and Accident Events

V. Charissis, S. Arafat, M. Patera (UK), C. Christomanos (Greece), and W. Chan (UK)


Artificial intelligence, modelling behaviour, driving simulation, HMI


A synthetic driving simulation environment requires a fleet of intelligent vehicles which can perform manoeuvring decisions in real-time and mingle orderly to form mixed traffic environments. The development of a driving simulator was part of a bigger project that evaluated a prototype Head-Up Display (HUD) interface. We therefore opted for a custom simulator that could effectively adapt to the needs of the given experiment. This paper presents the challenges we encountered during the implementation of realistic artificial intelligence (AI) in the traffic vehicles, which produced the analogous complexity driving events. A discussion follows regarding the design and modelling challenges of the robot drivers involved in the simulated accident scenarios. The paper discusses the outcomes of the user trials, suggests solutions to intriguing problems that derived from the experiment and reveals the phasma of our future work.

Important Links:

Go Back