Securing Interaction Between Human and Robot Using Expert Control System

摘要

Ensuring safety while using robots in their interaction with humans is one of the main principles of robotics. An expert control system for a robot would be based on the knowledge about such tools, which could help to solve applied problems in that field, where the robot is supposed to be used by humans. Such a system should also include expert, heuristic knowledge on communicative relations of the «human-robot» type. Those relations affect the selection, made by the robot of such problem-solving tools, which would be safe for humans. Thus, the knowledge on communicative relations determines the knowledge on problem-solving tools. Therefore, the application of an expert system for the purposes of robot control should provide a higher safety than a neural network system based on the procedure of machine learning. The goal of this research is to develop such a component of the expert control system, which would exclude the possibility of a robot causing health and life damage to a human. The method of achieving this goal consists of integrating a phenomenon of self-organization, inherent to a collective control subject within intersubjective «human-robot» relations, into production rules of the robot knowledge base. The knowledge base of the expert robot-control system should include the knowledge about equivalence of communicative relations between humans and robots. The research results in such a concept model of that component in the knowledge base of the expert robot-control system which would enable both the mechanism of feedback between a subject and an object of the control and the safety of managing such a complex intellectual technical system as a robot. The study is interdisciplinary by its nature and unites several approaches from the fields of cybernetics, synergetics, artificial intelligence, the logic of relations and gnosiology. The research concludes that an expert robot-control system, based on self-organization, provides safe cooperation between robots and humans.