Public policy makers have high expectations from automation in driving. The autonomous vehicle is expected to perform as a minimum the driving tasks of an alert non-aggressive driver. To go beyond these requirements, automation functions are to be designed to extend human driver capability in the form of processing large volume of data on driving environment in support of decision-making. This paper defines the role of Bayesian artificial intelligence (AI) in the development of car-following and lane change models for assisted and autonomous driving. Variables and model forms are covered. Driving simulator data in association with models enable illustrations on how new Bayesian AI-based models in driving assistance and automation system design can overcome safety problems caused by deficient perception-reaction performance. Finally, conclusions are presented on the roles of the Bayesian AI in adding cognitive capability to car-following and lane change models to be used for automation in driving.

Additional Metadata
Keywords Automation in driving, Bayesian artificial intelligence, Car-following model, Human-Factors, Lane-change model, Safety factors
Persistent URL dx.doi.org/10.1007/978-3-319-93885-1_28
Series Advances in Intelligent Systems and Computing
Citation
Khan, A. (2019). Role of Artificial Intelligence in Car-Following and Lane Change Models for Autonomous Driving. In Advances in Intelligent Systems and Computing. doi:10.1007/978-3-319-93885-1_28