Improving autonomous vehicle safety—Using acoustic source localization to influence the decision making capabilities of an autonomous vehicle

Abstract

The manner in which autonomous vehicles (AVs) and other vehicles will coexist, and communicate with one another, is still unclear, especially during the prolonged period of mixed vehicles sharing the road. During this time, it will be important that AVs are equipped to respond to acoustic alerts in urban environments. Consider the case of an emergency vehicle in a city. In this situation, a visual perception system alone will not suffice, as it requires a direct view of the source. The goal of this ongoing project is to define a new set of audio-visual detection and localization tools to identify the location of an unexpected rapidly approaching emergency vehicle. In particular, this paper focuses on the localization of a sound source in a complex environment by combining a ray tracing approach with a direction-of-arrival algorithm. This algorithm reports a number of source directions, arising from multiple reflections in the environment. Results are combined with a three-dimensional map, acquired live from a mobile robot equipped with digital sensors including LiDAR, and a reverse ray tracing approach is used to triangulate the likely position of the source.