Lifelike Robotic Collaboration requires Lifelike Information Integration

摘要

This paper is directed towards the definition of a systemic characteristic suitable for the 'intelligent' core of a cooperative robot. A major issue in developing cooperative systems is the nature and degree of their autonomy. It is rare (rather, impossible) that a communicated set of instructions for carrying out a task will be exhaustively complete, and with the passage of time the instruction set and the task's requirements often diverge. Neither complete dependence on control by a human 'master', nor complete autonomy from control is suitable or desirable within such a context, but it is as yet unclear how a satisfactory context-dependent intermediate modus can be developed which is conducive to cooperation while not stifling any preexistent or nascent capability for creative problem-solving. We believe that detailed examination of a number of more general aspects of system operation should predate attempts at defining performance measures for 'intelligent' systems per se.