An Interface for Switching between Autonomous and Manual Driving Using Location Information

Abstract

This paper introduces an interface for switching between autonomous and manual driving. A robot’s position and orientation are estimated using location information such as GNSS. Omnidirectional images are converted to equirectangular images. An augmented reality (AR) system superimposes location information onto it. Bird’s eye view images are generated using the robot’s position and orientation and location information. A self-confidence of the autonomous system is calculated using a covariance of the sensor. We conducted an experiment to investigate relationship between proposed location information display and usability to switch between autonomous and manual driving. The result described that proposed location information display improves usability.