132273Driver experience based learning system for autonomous carsClosedFeasibility StudiesInnovate UKThe sensor and recording mechanisms utilised in telematics insurance systems record driving data to analyse behaviour and risk according to individual mobility and external risk factors. The aim of this project is to investigate whether such data, following anonymisation could also be used to inform automated driving algorithms in autonomous cars; i.e. by influencing automated behaviour when considering mass behaviour of other vehicles. Correct decision making by future autonomous cars will crucially make judgements in various traffic situations and their trust and comprehension of other drivers typical background mobility data should likely be considered. This project aims to investigate the use of mass manual driver data to: 1) improve autonomous vehicle decision making capabilities and 2) Facilitate validation means of autonomous vehicle modelling and testing. This project aim is especially important in mixed transition environments (i.e. containing a mixture of autonomous and manual driving vehicles) to provide understanding of how autonomous vehicles ‘fit-in’ based upon normalised road and driver behaviour in order to both minimise and quantify risk.