Cliff-sensor-based Low-level Obstacle Detection for a Wheeled Robot in an Indoor Environment

A ramp and uneven ground formed by a low-level obstacle – whose height is too low from the ground – often stalls a robot’s navigation in an indoor environment. Few-centimeter differences between a low-level obstacle and a low-level non-obstacle are very difficult to be precisely measured in a constant distance at a mobile robot. In this paper, a wheeled mobile robot thus makes physical contact onto a low-level object, in order to measure such subtle differences. We use one or more cliff sensors typically in place for a mobile robot in order to avoid a drop-off. A wheeled robot climbs over a low-level object, classifies an obstacle vs. a non-obstacle using a cliff sensor’s timeseries data, and rapidly backs up before getting stuck onto an obstacle. While adopting a simplified deep-learning architecture, we suggest a rapid and accurate obstacle detection technique in real-time. We implemented our technique on an embedded robot platform of LG Hom-Bot. The supplementary video on the physical robot experiment can be accessed at https://youtu.be/yK57S857_II.