Probabilistic mobile robot with quantum decision-making

This work addresses informational-based methods for decision-making in Hilbert space, widely known as quantum decision-making, and their application to the control of mobile robot. We consider a mobile robot, whose objective is to find objects in a changing environment. We compare the robot performance under both a classical and a quantum decision making (control) approaches. Both approaches are driven by similar information-theory decision criteria. The robot behavior is physically studied by using an NXT Lego Robot at the CIM lab in Tel Aviv University.

1. INTRODUCTION

As indicated by Tsetlin and Fu, probabilistic automatons can be used for the modeling of automatic control systems, acting either in deterministic or stochastic environments. In particular, such models are suitable for the control of mobile robots, and give a good framework both for planning of robot's tasks and for programming its actions [9].