TY - GEN
T1 - UW-MARL
T2 - 2019 International Conference on Underwater Networks and Systems, WUWNET 2019
AU - Rahmati, Mehdi
AU - Nadeem, Mohammad
AU - Sadhu, Vidyasagar
AU - Pompili, Dario
N1 - Funding Information:
Acknowledgment: This work was supported by the NSF CPS Award No. 1739315. We thank Agam Modasiya and Karun Kanda (Rutgers MAE and CS students) for their help with the experiments.
Publisher Copyright:
© 2019 ACM.
PY - 2019/10/23
Y1 - 2019/10/23
N2 - Near-real-time water-quality monitoring in uncertain environments such as rivers, lakes, and water reservoirs of different variables is critical to protect the aquatic life and to prevent further propagation of the potential pollution in the water. In order to measure the physical values in a region of interest, adaptive sampling is helpful as an energy-and time-efficient technique since an exhaustive search of an area is not feasible with a single vehicle. We propose an adaptive sampling algorithm using multiple autonomous vehicles, which are well-trained, as agents, in a Multi-Agent Reinforcement Learning (MARL) framework to make efficient sequence of decisions on the adaptive sampling procedure. The proposed solution is evaluated using experimental data, which is fed into a simulation framework. Experiments were conducted in the Raritan River, Somerset and in Carnegie Lake, Princeton, NJ during July 2019.
AB - Near-real-time water-quality monitoring in uncertain environments such as rivers, lakes, and water reservoirs of different variables is critical to protect the aquatic life and to prevent further propagation of the potential pollution in the water. In order to measure the physical values in a region of interest, adaptive sampling is helpful as an energy-and time-efficient technique since an exhaustive search of an area is not feasible with a single vehicle. We propose an adaptive sampling algorithm using multiple autonomous vehicles, which are well-trained, as agents, in a Multi-Agent Reinforcement Learning (MARL) framework to make efficient sequence of decisions on the adaptive sampling procedure. The proposed solution is evaluated using experimental data, which is fed into a simulation framework. Experiments were conducted in the Raritan River, Somerset and in Carnegie Lake, Princeton, NJ during July 2019.
KW - Autonomous underwater vehicles
KW - Field experiments
KW - Multi-agent reinforcement learning
KW - Underwater adaptive sampling
UR - http://www.scopus.com/inward/record.url?scp=85081066423&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85081066423&partnerID=8YFLogxK
U2 - 10.1145/3366486.3366533
DO - 10.1145/3366486.3366533
M3 - Conference contribution
AN - SCOPUS:85081066423
T3 - ACM International Conference Proceeding Series
BT - Proceedings of the International Conference on Underwater Networks and Systems, WUWNET 2019
PB - Association for Computing Machinery
Y2 - 23 October 2019 through 25 October 2019
ER -