Real-Time and Near-Optimal Decision-Making in Multi-Robot Systems for Ocean Science
Date:
Talk Information
Date: Monday, September 8, 2025
Venue: University of Georgia – Skidaway Institute of Oceanography (SkIO)
Video Recording
You can watch the seminar at the link below:
🔗 Recording:
https://skio.webex.com/skio/ldr.php?RCID=98f4ec8fa70451a826b6140afe074945
🔑 Password: mYWuJva3
(Video hosted on Webex by UGA SkIO.)
Abstract
Multi-agent robotic systems—including fixed sensors, autonomous surface/underwater vehicles, and aerial drones—play an increasingly important role in environmental monitoring, maritime security, and resource management across dynamic and uncertain ocean environments.
This talk presents stochastic modeling approaches and heuristic optimization techniques that enable autonomous systems to operate in real time and deliver near-optimal decisions under uncertainty. Applications include optimizing sensor networks for detecting rare or transient phenomena, coordinating heterogeneous robot fleets for large-area monitoring, and integrating environmental data streams for adaptive mission planning.
The seminar highlights recent advances from projects supported by the Office of Naval Research, demonstrating how probabilistic methods and decision-making frameworks can be adapted to the complexities of marine environments. Additional examples illustrate how these approaches support harmful algal bloom detection, vessel tracking, and climate-driven habitat monitoring.
Description
This seminar introduces the algorithmic, probabilistic, and operational principles behind real-time decision-making in multi-robot systems. Examples emphasize interdisciplinary collaborations between robotics, ocean science, and defense applications, showing how autonomous systems can support emerging challenges in marine environments.