Mutualistic Interactions in Heterogeneous Multi-Robot Systems
Abstract
In multi-robot systems, having robots with different capabilities can be highly beneficial as it allows for the robot team to be abl,e to participate in a larger class of missions, as sensing, actuation, computation, or communication capabilities can be spread acro,ss different team-mates, thereby supporting the deployment of a potentially significantly larger number of distinct capabilities. Th,is proposal formalizes the idea that it should be possible to make the whole larger than the sum of its parts in multi-agent robotic,s. For that to happen, it is no longer enough to consider capabilities in isolation, but rather how they interact to form brand new,capabilities. In ecology, such types of arrangements are known as a mutualistic interactions, or mutualisms, involving the exchange,of resources or services between two different species. Each species involved in a mutualism receives a benefit from the interaction,, and what this proposal envisions is a new framework for harnessing and encoding the idea of mutualism in a multi-agent robotics se,tting in order to go beyond the standard approach to heterogeneity by having brand new capabilities arise when combining agents with, different individual capabilities. To that end, this multidisciplinary proposal brings together a roboticist with an ecologist. The, two particular topics that will be pursed to achieve this ambitious and multidisciplinary agenda are: (1) Characterization of Mutua,listic Interactions in Ecology: To be able to leverage the types of mutualistic interactions observed in nature, the scope and scale, of such interactions need to be compiled and modeled in a unified and computationally actionable manner. This will constitute Resea,rch Area 1 and will, in particular, focus on developing models of mutualisms as co-evolutionary processes, a robotics-relevant taxon,omy of mutualisms, and an investigation of the conditions under which mutualisms break down and fail to be effective.(2) Leveraging,Mutualistic Interactions in Robotics: A theory will be developed for how to describe combined multi-robot capabilities in such a way, that brand new capabilities result from the deployment of heterogeneous agents. In particular, Research Area 2 will investigate how, to formally understand how mutualisms can be leveraged in the broad contexts of coverage control and task assignment problems for h,eterogeneous multi-robot teams.Approved for Public Release
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 05, 2022
- Source ID
- N000142212625
Entities
People
- Magnus Egerstedt
Organizations
- Office of Naval Research
- United States Navy
- University of California, Irvine