Satisficing Action Selection

It is argued that for real-world applications action selection should be satisficing, i.e. merely `good enough' rather than optimal. It is then demonstrated that multiple objective decision theory provides a suitable framework for formulating action selection mechanisms that are satisficing. A set of experiments demonstrate the potential advantages of the proposed action selection mechanisms. Keywords: Action selection, behavior-based control, multiple objective decision making, Pareto-optimality, satisficing actions 1. INTRODUCTION Operation of robots in unknown settings poses complex control problems which require solutions that guarantee a suitable trade-off between a multitude of (potentially) conflicting task objectives. For instance, one action (or set of control parameters) which is optimal with respect to one particular objective (e.g., speed) might render unacceptable with respect to other objectives (e.g., safety). Thus a major issue in the design of systems for control of..