A Conceptual Characterization of Autonomy in the Philosophy of Robotics

The concept of autonomy is crucial for the theoretical characterization of robots and, more in general, complex technological artifacts. The aim of this paper is to provide a conceptual and logical framework in which it is possible to define two concepts of autonomy: autonomy of performance and autonomy of process. The analysis is carried out exploiting the logical resources of the counterfactual semantics – developed by Lewis’ and Stalnaker’s seminal works – and branching structures of the possible courses of actions. It allows to differentiate the autonomy of a robotic arm from the autonomy of a highly complex machine- like rovers for the explorations of the planets. The clarification of the concept of autonomy is, moreover, an essential precondition for the investigation concerning the ethics of artificial agents.