GoalSwarm: Multi-UAV Semantic Coordination for Open-Vocabulary Object Navigation

TL;DR

GoalSwarm enables decentralized multi-UAV teams to perform open-vocabulary object navigation efficiently by constructing shared semantic maps, integrating foundation models, and coordinating exploration without heavy onboard computation.

Contribution

It introduces a novel decentralized framework that combines lightweight semantic mapping, foundation model integration, and multi-agent coordination for open-vocabulary navigation.

Findings

Achieves robust zero-shot object navigation in unknown environments.

Reduces onboard computational load with 2D semantic mapping.

Demonstrates effective multi-UAV coordination for exploration.

Abstract

Cooperative visual semantic navigation is a foundational capability for aerial robot teams operating in unknown environments. However, achieving robust open-vocabulary object-goal navigation remains challenging due to the computational constraints of deploying heavy perception models onboard and the complexity of decentralized multi-agent coordination. We present GoalSwarm, a fully decentralized multi-UAV framework for zero-shot semantic object-goal navigation. Each UAV collaboratively constructs a shared, lightweight 2D top-down semantic occupancy map by projecting depth observations from aerial vantage points, eliminating the computational burden of full 3D representations while preserving essential geometric and semantic structure. The core contributions of GoalSwarm are threefold: (1) integration of zero-shot foundation model -- SAM3 for open vocabulary detection and pixel-level segmentation, enabling open-vocabulary target identification without task-specific training; (2) a Bayesian Value Map that fuses multi-viewpoint detection confidences into a per-pixel goal-relevance distribution, enabling informed frontier scoring via Upper Confidence Bound (UCB) exploration; and (3) a decentralized coordination strategy combining semantic frontier extraction, cost-utility bidding with geodesic path costs, and spatial separation penalties to minimize redundant exploration across the swarm.