Architectures for Shared Control of Vehicle Steering

Abstract

The allocation of authority between a human driver and an automation system designed to share control with the driver must be managed carefully to ensure safety and driver satisfaction. A driver can tighten grip on a steering wheel, co-contract arm muscles, and change posture to increase authority relative to an automation system that has a finite mechanical impedance. In this paper we develop a system model outfitted with the features required to describe modulating impedance as a means to dynamically allocate authority. We adopt a standard modeling framework, but focus on the description of the driving point impedance of both the driver and the automation system. Haptic feedback from the steering wheel functions both to cue the driver but also to couple the driver biomechanics to the steering and automation system. We also describe a platform designed to support experiments in control sharing with dynamic allocation of authority between two humans, one of them acting as a stand-in for the automation system.