Using Knowledge Ontologies and Neural Networks to Control Service - Oriented Robots

In recent years, researchers have been building home robots able to interact and work with people. However, due to the complexity and diversity of the environments for developing robots, it is difficult to share and reuse robot code created by different providers. In this work we present a servicebased approach that exploits the standard web interface to develop reusable robotic services. Our approach includes knowledge ontology and neural network learning strategies for robot control. In addition, several service functions, including service discovery, selection, and composition have been developed for operating the services. The proposed approach has been implemented and evaluated, and the results show that it can be used to build robotic services successfully.