Autonomous field mapping and operation with proximity sensing

This paper describes a method of autonomous field mapping and operation with proximity sensing. Localization and mapping are important functions for mobile robots to enhance the efficiency of operations in the work field. Conventionally, the GPS-based localization and LRF-based localization are popular for mobile robots working outdoor and indoor fields, respectively. However, the GPS is not enough in accuracy for operation in small area, and D-GPS costs too high for commercial implementations. The LRF-based method functions well in indoor environments, but is difficult for localization in the outdoor field without architectural structures to scan. In this paper, we propose a novel method of field sensing using proximity sensors. The proximity sensor allows to identify a boundary and landmarks embedded in the field, and they can be easily installed into commercial products in low cost. We developed some algorithms of autonomous mapping and operation behavior based on proximity sensing, and examined the effectiveness with a real mobile robot in a test field.