| NSF Org: |
IIS Division of Information & Intelligent Systems |
| Recipient: |
|
| Initial Amendment Date: | June 30, 2016 |
| Latest Amendment Date: | June 30, 2016 |
| Award Number: | 1617838 |
| Award Instrument: | Standard Grant |
| Program Manager: |
Erion Plaku
eplaku@nsf.gov (703)292-0000 IIS Division of Information & Intelligent Systems CSE Directorate for Computer and Information Science and Engineering |
| Start Date: | October 1, 2016 |
| End Date: | September 30, 2020 (Estimated) |
| Total Intended Award Amount: | $192,557.00 |
| Total Awarded Amount to Date: | $192,557.00 |
| Funds Obligated to Date: |
|
| History of Investigator: |
|
| Recipient Sponsored Research Office: |
6729 NW 39TH EXPY BETHANY OK US 73008-2694 (405)491-6329 |
| Sponsor Congressional District: |
|
| Primary Place of Performance: |
OK US 73008-2605 |
| Primary Place of
Performance Congressional District: |
|
| Unique Entity Identifier (UEI): |
|
| Parent UEI: |
|
| NSF Program(s): | Robust Intelligence |
| Primary Program Source: |
|
| Program Reference Code(s): |
|
| Program Element Code(s): |
|
| Award Agency Code: | 4900 |
| Fund Agency Code: | 4900 |
| Assistance Listing Number(s): | 47.070 |
ABSTRACT
The ability to use teams of robots in real-world tasks, such as exploration, reconnaissance and search and rescue, depends on their ability to effectively cooperate in complex and dynamic environments. Unfortunately, determining the most effective team size for a given task depends on a variety of factors, often requires information not readily available, and can be computationally impractical. Observations from nature show that some animal societies exhibit frequent splits (fission) and merges (fusion) of subgroups without the coordination overhead frequently found in multi-robot systems. The goal of this research project is to use these insights from fission-fusion societies in nature as inspiration to implement similar behavior in multi-robot systems.
This project will investigate the motivations and mechanisms that contribute to the fission of a team of robots into smaller groups and the fusion of smaller groups into a larger group, into sizes appropriate to a given task. Potential biological and environmental factors that contribute to individual decisions that result in the fission and fusion of groups and that have artificial analogues relevant to multi-robot systems will be identified using agent-based simulations. An emphasis will be placed on factors that contribute to an adaptive task-specific preferred group size. Once identified, these factors will be implemented on physical robots to evaluate the performance of the decision-making system within the constraints of physical robots. The decision-making system used will be implemented using an adaptive fuzzy behavior hierarchy and a collective decision-making model developed using observations of natural systems.
PROJECT OUTCOMES REPORT
Disclaimer
This Project Outcomes Report for the General Public is displayed verbatim as submitted by the Principal Investigator (PI) for this award. Any opinions, findings, and conclusions or recommendations expressed in this Report are those of the PI and do not necessarily reflect the views of the National Science Foundation; NSF has not approved or endorsed its content.
The ability to use teams of autonomous robots in interesting, real-world tasks such as exploration, reconnaissance, and search and rescue depends on their ability to effectively and efficiently cooperate in complex and dynamic environments. For some tasks, a small number of large groups is the best option, while other tasks may require a large number of small groups. Although the optimal configuration for some tasks may be determined, these solutions are frequently unable to adapt to dynamic environments, rapid changes in tasks, or situations in which team membership frequently changes. While this is a significant challenge for multi-robot systems (MRSs), natural systems exhibit frequent changes in subgroup composition and number in response to factors such as resource availability, conflicts of interest, and predation risk. This fission-fusion process allows natural systems to adapt just as an MRS should. The goal of this project was to gain inspiration from natural systems exhibiting fission-fusion behaviors and design MRSs that can adapt subgroup size and number dynamically, depending on the current task.
In order to meet this goal, we investigated the environmental factors that promote fission-fusion events in natural systems, especially those that have relevant analogues in artificial systems such as an MRS. Once we identified potential factors, we designed control systems that made use of the promising factors and evaluated them in simulated environments with a group of simulated robots. Special attention was paid to using only information that could realistically be sensed by a physical robot.
We made several important steps towards achieving our goal in this project. First, we were able to demonstrate a design for an MRS that promotes fission-fusion behaviors without defining explicit groups. Preliminary progress was also made on extending the MRS so that fission-fusion events can be produced as a byproduct of environmental changes and not explicit incentives. To aid in the development of these MRSs, we also made progress on developing an approach for evolving comprehensible fuzzy rulesets. The intent here was to create MRS systems that could be analyzed by researchers from many backgrounds that do not traditionally have experience in analyzing complex control systems. These steps, however, are just the beginning in this line of research. We will continue to refine the design of these systems and approaches and extend them to more general applications in the coming years.
Throughout the entire project, undergraduate research assistants made significant contributions. They were involved in everything from determining the direction of various aspects of the research to the implementation and analysis of the experiments.
Last Modified: 12/30/2020
Modified by: Brent E Eskridge
Please report errors in award information by writing to: awardsearch@nsf.gov.
