Supervisory control theory applied to swarm robotics

On

Abstract

At present, most of the source code controlling swarm robotic systems is developed in an ad-hoc manner. This can make it difficult to maintain these systems and to guarantee that they will accomplish the desired behaviour.

Formal approaches can help to solve these issues. However, they do not usually guarantee that the final source code will match the modelled specification.

To address this problem, our research explores the application of formal approaches to both synthesise high-level controllers and automatically generate control software for a swarm of robots.

The formal approach used in this paper is the supervisory control theory. The approach is successfully validated in two experiments using up to 42 Kilobot robots and up to 26 e-puck robots.


Software tool

The software tool used is Nadzoru.


Resources

View and download the free behaviour models and control specifications.

View and download the source code.


Videos

Five times real time speed.

Orbit e-puck: Trials 1-10
Trial 1
Trial 2
Trial 3
Trial 4
Trial 5
Trial 6
Trial 7
Trial 8
Trial 9
Trial 10
Orbit Kilobot: Trials 1-10
Trial 1
Trial 2
Trial 3
Trial 4
Trial 5
Trial 6
Trial 7
Trial 8
Trial 9
Trial 10
Segregation e-puck: Trials 1-10
Trial 1
Trial 2
Trial 3
Trial 4
Trial 5
Trial 6
Trial 7
Trial 8
Trial 9
Trial 10
Segregation Kilobot: Trials 1-10
Trial 1
Trial 2
Trial 3
Trial 4
Trial 5
Trial 6
Trial 7
Trial 8
Trial 9
Trial 10
Extra

Orbit 2 e-pucks

Orbit 3 e-pucks

Orbit 4 e-pucks

Orbit 2 Kilobots

Orbit 3 Kilobots (a)

Orbit 3 Kilobots (b)

Project updates

Natural Robotics Lab: investigating robotic systems inspired by nature, and robotic models of natural systems.

A global reputation

Sheffield is a research university with a global reputation for excellence. We're a member of the Russell Group: one of the 24 leading UK universities for research and teaching.