Distracted driving or driving in a tired state causes many road accidents. One of the approaches to address this problem is to use an automated solution that detects driver state and executes some actions, aimed on maintaining driver’s attention focus. Personalized intervention systems are actively developing in the field of medicine, however, in the field of transport, no similar studies have been found. This paper examines the idea of using reinforcement learning to generate personalized driver interventions. The contribution of the paper is twofold. First, it proposes a conceptual model and general reinforcement learning formulation of the problem. Second, it describes a driver simulation model that can be used to train the personalized intervention policy. Experimental study shows, that the proposed formulation allows one to train a personalized policy that can be used to effectively maintain the desired state of the vehicle driver. The implementation of such method in a driver monitoring system can be a very promising direction and help to reduce the number of road accidents for the above reasons.
Maintaining Vehicle Driver’s State Using Personalized Interventions
2022-04-27
432153 byte
Conference paper
Electronic Resource
English
Maintaining Driver's Vigilance for Preventive Safety
British Library Conference Proceedings | 1995
|DRIVER'S STATE ESTIMATION APPARATUS AND DRIVER'S STATE ESTIMATION METHOD
European Patent Office | 2015
|