Redirecting to original paper in 30 seconds...

Click below to go immediately or wait for automatic redirect

arxiv_ml 85% Match Research Paper Robotics Engineers,Control Systems Researchers,AI Researchers in Multi-Agent Systems 20 hours ago

A Spatially Informed Gaussian Process UCB Method for Decentralized Coverage Control

reinforcement-learning › multi-agent
📄 Abstract

Abstract: We present a novel decentralized algorithm for coverage control in unknown spatial environments modeled by Gaussian Processes (GPs). To trade-off between exploration and exploitation, each agent autonomously determines its trajectory by minimizing a local cost function. Inspired by the GP-UCB (Upper Confidence Bound for GPs) acquisition function, the proposed cost combines the expected locational cost with a variance-based exploration term, guiding agents toward regions that are both high in predicted density and model uncertainty. Compared to previous work, our algorithm operates in a fully decentralized fashion, relying only on local observations and communication with neighboring agents. In particular, agents periodically update their inducing points using a greedy selection strategy, enabling scalable online GP updates. We demonstrate the effectiveness of our algorithm in simulation.

Key Contributions

Introduces a novel decentralized algorithm for coverage control in unknown spatial environments using Gaussian Processes. It enables agents to autonomously determine trajectories by minimizing a local cost function that balances expected locational cost with variance-based exploration, inspired by GP-UCB. The algorithm operates fully decentralized, relying on local observations and neighbor communication, and uses a greedy selection strategy for scalable online GP updates.

Business Value

Enables efficient and autonomous operation of multi-agent systems for tasks like environmental surveying, search and rescue, or precision agriculture, reducing the need for constant human supervision and improving coverage efficiency.