S. Mahi, K. Nam, and C. Crick. Proceedings of the 18th International Conference on Autonomous Agents and MultiAgent Systems, page 1359–1367. Richland, SC, International Foundation for Autonomous Agents and Multiagent Systems, (2019)
Abstract
We introduce a novel, scalable, distributed decision-making algorithm using factor graphs and the sum product algorithm to control the coordination of a heterogeneous multi-robot team in exploration tasks. In addition, our algorithm supports seamless participation of human operators at arbitrary levels of interaction. We present experimental results performed using both simulated and actual teams of unmanned aerial systems (UAS). Our experiments demonstrate effective exploration while facilitating human participation with the team. At the same time, we show how robots with differing capabilities coordinate their behaviors effectively to leverage each other's individual strengths, without having to explicitly account for every possible joint behavior during system design. We demonstrate our algorithm's suitability for tasks such as weather data collection using a heterogeneous robot team consisting of fixed- and rotary-wing UAS. In particular, during 60 flight hours of real-world experiments collecting weather data, we show that robots using our algorithm were about seven times more efficient at exploring their environment than similar systems which flew preplanned flight profiles. One of our primary contributions is to demonstrate coordinated autonomous control and decision-making among robots operating in very different flight regimes.
%0 Conference Paper
%1 mahi2019humanrobot
%A Mahi, S M Al
%A Nam, Kyungho
%A Crick, Christopher
%B Proceedings of the 18th International Conference on Autonomous Agents and MultiAgent Systems
%C Richland, SC
%D 2019
%E Elkind, Edith
%E Veloso, Manuela
%E Agmon, Noa
%E Taylor, Matthew E.
%I International Foundation for Autonomous Agents and Multiagent Systems
%K agent human hybrid interaction mas multi robot system team
%P 1359–1367
%T Distributed Heterogeneous Robot-Human Teams: Robotics Track
%X We introduce a novel, scalable, distributed decision-making algorithm using factor graphs and the sum product algorithm to control the coordination of a heterogeneous multi-robot team in exploration tasks. In addition, our algorithm supports seamless participation of human operators at arbitrary levels of interaction. We present experimental results performed using both simulated and actual teams of unmanned aerial systems (UAS). Our experiments demonstrate effective exploration while facilitating human participation with the team. At the same time, we show how robots with differing capabilities coordinate their behaviors effectively to leverage each other's individual strengths, without having to explicitly account for every possible joint behavior during system design. We demonstrate our algorithm's suitability for tasks such as weather data collection using a heterogeneous robot team consisting of fixed- and rotary-wing UAS. In particular, during 60 flight hours of real-world experiments collecting weather data, we show that robots using our algorithm were about seven times more efficient at exploring their environment than similar systems which flew preplanned flight profiles. One of our primary contributions is to demonstrate coordinated autonomous control and decision-making among robots operating in very different flight regimes.
%@ 9781450363099
@inproceedings{mahi2019humanrobot,
abstract = {We introduce a novel, scalable, distributed decision-making algorithm using factor graphs and the sum product algorithm to control the coordination of a heterogeneous multi-robot team in exploration tasks. In addition, our algorithm supports seamless participation of human operators at arbitrary levels of interaction. We present experimental results performed using both simulated and actual teams of unmanned aerial systems (UAS). Our experiments demonstrate effective exploration while facilitating human participation with the team. At the same time, we show how robots with differing capabilities coordinate their behaviors effectively to leverage each other's individual strengths, without having to explicitly account for every possible joint behavior during system design. We demonstrate our algorithm's suitability for tasks such as weather data collection using a heterogeneous robot team consisting of fixed- and rotary-wing UAS. In particular, during 60 flight hours of real-world experiments collecting weather data, we show that robots using our algorithm were about seven times more efficient at exploring their environment than similar systems which flew preplanned flight profiles. One of our primary contributions is to demonstrate coordinated autonomous control and decision-making among robots operating in very different flight regimes.},
added-at = {2020-04-01T13:24:45.000+0200},
address = {Richland, SC},
author = {Mahi, S M Al and Nam, Kyungho and Crick, Christopher},
biburl = {https://www.bibsonomy.org/bibtex/20a3c1855bcb4a891158ebef1416f3c17/porta},
booktitle = {Proceedings of the 18th International Conference on Autonomous Agents and MultiAgent Systems},
editor = {Elkind, Edith and Veloso, Manuela and Agmon, Noa and Taylor, Matthew E.},
interhash = {f1a2fdb2adbb7c576513cad2822d3383},
intrahash = {0a3c1855bcb4a891158ebef1416f3c17},
isbn = {9781450363099},
keywords = {agent human hybrid interaction mas multi robot system team},
location = {Montreal QC, Canada},
numpages = {9},
pages = {1359–1367},
publisher = {International Foundation for Autonomous Agents and Multiagent Systems},
series = {AAMAS ’19},
timestamp = {2020-06-14T15:15:59.000+0200},
title = {Distributed Heterogeneous Robot-Human Teams: Robotics Track},
year = 2019
}