Within this contribution we present the sensory-actuatory map SAM, as an alternative approach to build topological maps. The structure of SAM is presented together with a description of how to implement it. Seven core capabilities that have been identified to be necessary for the SAM approach are defined, and explained. In addition, the successful implementation of the SAM approach onto a real six-wheeled autonomous robot system is presented. The algorithms applied for each of the seven core capabilities have been adapted for the special needs of a robot platform with limited, and sparse sensory capabilities. Modules for sensory item identification, storage, and recognition have been implemented based on processing of sensory time series. The application of SAM for exploring and mapping of a test environment, and the advantages and limitations of the developed approach are presented and discussed.
展开▼
