University of Sheffield
Force-Aware Self-Assembly of Bridges by a Swarm of Autonomous Robots
Abstract
dc:description.abstractSelf-assembling modular robotic systems comprise multiple agents which form connections with each other to create different arrangements to suit their current task. Existing control algorithms for these systems typically do not consider the forces in the connections between agents, potentially leading to damage to the robots due to unacceptably high forces. This thesis considers how force-aware methods can be used to guide the agents into self-assembling structures that will not break. As an example scenario where force-aware control shows great potential, the self-assembly of a bridge across a gap in the terrain is considered. This would enable a group of robots to explore otherwise unreachable environments, but requires careful consideration to ensure the bridge does not collapse under its own weight. This process is split into three stages, and separate algorithms are developed to allow each stage to be completed: each algorithm runs in a distributed manner across each agent, and responds to measurements of forces made by the agents to produce strong structures that are not designed in advance by a human. A cantilever is first constructed from one side of the gap until the other is reached, whereupon the structure is optimised by removing agents that are no longer providing support. Finally, the bridge is deconstructed when it is no longer required. The algorithms are first verified in simulation, then a novel hardware platform is developed to demonstrate their applicability in real life. The cantilevers and bridges built are found to be close to the optimal with respect to the number of agents required to build a stable structure of a given length. The forces within the structures can be controlled by choosing suitable allowable limits. This work hopes to inspire future researchers to explore how force-aware methods can be applied to other problems in self-assembling modular robotics.
Degree
thesis:*- Name dc:type.qualificationname
- Ph.D
- Level dc:type.qualificationlevel
- doctoral
- Grantor dc:publisher.institution
- University of Sheffield
- Year dc:date.issued
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Bray, Edward Nicholas
- Advisors dc:contributor.advisor
-
- Gross, Roderich
- Damian, Dana
- ORCID person:identifier.orcid
Rights
- Language dc:language.dcterms_iso639_2
- eng
Identifiers
dc:identifier.*- Dc Identifier Grantnumber
- 199569285