Supervisory Coordination of Robotic Fiber Positioners in Multi-Object Spectrographs

TL;DR

This paper applies supervisory control theory to coordinate robotic fiber positioners in multi-object spectrographs, modeling their behaviors as discrete-event systems and synthesizing supervisors for complete coordination.

Contribution

It introduces a novel supervisory control approach for robotic fiber positioner coordination, including a new algorithm based on backtracking forcibility for reconfiguration tasks.

Findings

Successfully models positioners as discrete-event systems

Synthesizes coordination supervisors for positioner groups

Demonstrates effectiveness with an illustrative example

Abstract

In this paper, we solve the complete coordination problem of robotic fiber positioners using supervisory control theory. In particular, we model positioners and their behavioral specifications as discrete-event systems by the discretization of their motion spaces. We synthesize a coordination supervisor associated with a specific set of positioners. In particular, the coordination supervisor includes the solutions to the complete coordination problem of its corresponding positioners. Then, we use the backtracking forcibility technique of supervisory control theory to present an algorithm based on a completeness condition to solve the coordination problem similar to a reconfiguration problem. We illustrate the functionality of our method using an example.