A Mobile Homeostat with Three Degrees of Freedom

Abstract

In the foundational cybernetics text, Design for a Brain, W. Ross Ashby introduces an adaptive system called the homeostat, and speculates about the possibility of creating a mobile homeostat "with its critical states set so that it seeks situations of high illumination." Simulations demonstrate the viability of using the classic homeostat architecture to control a mobile robot demonstrating ultrastability in adapting to an environment where the goal is to stay within range of a single source of illumination. This paper explores a novel physical embodiment of Ashby's classic homeostat in a mobile robot with three degrees of freedom (2 translational, 1 rotational). The hypothesis, borne out by tests, is that the topological configuration of the robot as determined by simulation, will carry over into the physical robot.

Learning the Ashby Box: an experiment in second order cybernetic modeling

“…A fundamental difference between the well-studied Homeostat (Di Paolo, 2003;Asaro, 2008;Harvey, 2008;Cariani, 2009;Battle, 2015) and the NTM/observer architecture lies in the way goals are defined. Since in the general case, the observer's goals are relative to the current input state (thus there are no "essential variables" as such), the apparent control exerted by the goal-seeking observer on the NTM does not necessarily involve an absolute reduction in NTM variety.…”