Acoustic-based room discrimination for the navigation of autonomous mobile robots

摘要

This work investigates the possibility of using audible broadband acoustic signals for the navigation of autonomous mobile robots in unprepared, unknown, silent, static, indoor environments. Descriptive psychological experiments show that humans have astounding abilities in differentiating their relative position to walls and sensing the existence of even small objects like metal discs not bigger than a plate, solely based on acoustic information. As there exists no theory on the practical limits of discriminations spaces based on acoustics, an experimental setup is investigated that enables conclusions about the applicability of acoustics for navigation purposes. First, straight forward approach to a representation and metric for the acoustics of the surrounding is described and tested within the experimental setup.