Technology Category
Multiagent Systems: Agent CommunicationReasoning under Uncertainty: Sequential Decision MakingMachine Learning: Calibration & Uncertainty Quantification
Decentralized Gaussian Process Classification and an Application in Subsea Robotics
🤖 AI Summary
Acoustic communication in multi-AUV cooperative missions suffers from high uncertainty, low bandwidth, and difficulty in real-time environmental modeling. Method: This paper proposes a decentralized Gaussian process classification framework for online construction and updating of probabilistic underwater communication success maps. It introduces a rigorously derived data-sharing strategy that optimally selects and fuses sparse, asynchronous measurements across AUVs—without relying on a central node—to enhance mapping efficiency and generalization. The approach integrates distributed machine learning with real-time probabilistic modeling. Contribution/Results: Evaluated using real-world acoustic communication data from Virginia Tech’s 690-class AUVs, the method achieves a 23.6% improvement in communication success prediction accuracy over baseline approaches and demonstrates significantly enhanced system robustness, establishing a scalable environmental perception paradigm for resource-constrained underwater cooperative operations.
Technology Category
Application Category
📝 Abstract
Teams of cooperating autonomous underwater vehicles (AUVs) rely on acoustic communication for coordination, yet this communication medium is constrained by limited range, multi-path effects, and low bandwidth. One way to address the uncertainty associated with acoustic communication is to learn the communication environment in real-time. We address the challenge of a team of robots building a map of the probability of communication success from one location to another in real-time. This is a decentralized classification problem -- communication events are either successful or unsuccessful -- where AUVs share a subset of their communication measurements to build the map. The main contribution of this work is a rigorously derived data sharing policy that selects measurements to be shared among AUVs. We experimentally validate our proposed sharing policy using real acoustic communication data collected from teams of Virginia Tech 690 AUVs, demonstrating its effectiveness in underwater environments.
Problem
Research questions and friction points this paper is trying to address.
Learning real-time communication success probability maps underwater
Developing decentralized classification for acoustic communication events
Establishing data sharing policies for autonomous underwater vehicle teams
Innovation
Methods, ideas, or system contributions that make the work stand out.
Decentralized Gaussian Process Classification for AUV teams
Real-time acoustic communication success probability mapping
Data sharing policy selects measurements for AUV exchange
🔎 Similar Papers
No similar papers found.
