Robotic Swarm Control from Spatio-Temporal Specifications

TL;DR

This paper presents a method for controlling robotic swarms to achieve complex spatio-temporal patterns using formal logic and mixed integer linear programming, enabling precise and optimized swarm behaviors.

Contribution

It introduces a novel approach that encodes spatio-temporal specifications into MILP constraints for effective swarm control.

Findings

Successfully encodes complex spatio-temporal specs as MILP constraints.

Demonstrates optimized swarm trajectories satisfying specifications.

Includes a case study validating the approach.

Abstract

In this paper, we study the problem of controlling a two-dimensional robotic swarm with the purpose of achieving high level and complex spatio-temporal patterns. We use a rich spatio-temporal logic that is capable of describing a wide range of time varying and complex spatial configurations, and develop a method to encode such formal specifications as a set of mixed integer linear constraints, which are incorporated into a mixed integer linear programming problem. We plan trajectories for each individual robot such that the whole swarm satisfies the spatio-temporal requirements, while optimizing total robot movement and/or a metric that shows how strongly the swarm trajectory resembles given spatio-temporal behaviors. An illustrative case study is included.