Swarm Inspired Path-Constrained Motion Planning for Maintaining Communication Within Multiple Robot Systems
Year: 2024
Type: conference-paper
Abstract: The involvement of Multi-robot Systems (MRS) in a vast range of missions such as search and rescue, patrolling, exploration and mapping of environments has increased over the years. The working dynamics between MRS and human operators has also evolved to include more human-in-the-loop approaches where the MRS becomes an extension of the human operator(s) to accomplish complex tasks as human-robot teams. This advancement requires a higher frequency of information sharing among the MRS and with the human operator(s) to provide situational awareness so that more informed decisions can be made rapidly. Thus, maintaining communication between the various actors of the team is key to ensure mission success. This paper explores the dynamic challenge of swarm reconfiguration to maintain communication among a system of dynamic nodes — i.e. multiple robots or humans — navigating in an urban environment. Proposed is a hierarchical robotic system that utilises a swarm of Small Supporting Unmanned Ground Vehicles (SSUGVs) as mobile communication breadcrumbs. These SSUGVs dynamically reconfigure themselves through swarm-inspired, path-constrained motion planning to maintain communication within the main system during exploration. The presented solution is verified successfully with hybrid simulations — i.e. a combination of real experimental data with simulated components on a mapping exploration task. The system is studied based on three metrics namely the system connectivity that measures the network connectivity of the overall system, the system response that measures the response of the system to a changing environment, and the average distance travelled of the system that indicates the activity level of the system.
Authors: Jabez Kit Leong, Brandon J. Lawrence, Gim Song Soh
Institution: Singapore University of Technology and Design
Language: English
Subfield: Mechanical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4404310717