Quantitative Evaluation of the Exploration Strategies of a Mobile Robot

Abstract

This article describes an experimental investigation into the map-building and exploration capabilities of a mobile robot. Two types of map are used: a set of line and point features, and a grid-based free-space map. Potential features are ex tracted from sonar range readings and classed as "confirmed" if detected repeatedly. The free-space map is derived from the set of confirmed features. A distance-transform algorithm is then used to plan paths on this map. The confirmed features are used by a Kalman filter to estimate the robot's position relative to known objects. This research places exceptional stress on the need for practical experimentation and quantitative, statistical evaluation of the results. For this to be possible, it is essential to have a clearly defined measure of map quality. A novel metric is defined that predicts the effectiveness of the robot if it were to use the map to execute a set of test tasks. Exploration strategies are tested experimentally in a range of environments and starting positions. The results are eval uated and compared statistically. The tested strategies range from those that are totally reactive, such as wall-following, to those that use the developing map to focus attention on the un examined parts of the environment. The most promising results are observed from hybrid exploration strategies that combine the robustness of reactive navigation and the directive power of map-based strategies.