Swarming in the Wild: A Distributed Communication-less Lloyd-based Algorithm dealing with Uncertainties

2025-04-26

AI Summary

Multi-robot swarms operating in GNSS-denied, unknown complex environments (e.g., dense forests) must achieve safe, collaborative navigation and target convergence without inter-robot communication, global positioning, or prior maps. Method: We propose a fully distributed, onboard perception-driven control algorithm grounded in the Lloyd optimization framework. It integrates local sensing feedback, adaptive neighbor-constraint modeling, and finite-time convergence analysis, explicitly accounting for perception uncertainty and tracking errors while rigorously enforcing obstacle avoidance, inter-robot collision avoidance, and formation distance constraints. Contribution/Results: This work presents the first field demonstration of safe, communication-free, GPS-free swarm convergence in unstructured outdoor environments. Field experiments with multiple UAVs successfully accomplished target-area arrival in dense forest settings, satisfying all safety, proximity, and convergence requirements throughout the mission.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Search and Optimization: Distributed Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

In this work, we present a distributed algorithm for swarming in complex environments that operates with no communication, no a priori information about the environment, and using only onboard sensing and computation capabilities. We provide sufficient conditions to guarantee that each robot reaches its goal region in a finite time, avoiding collisions with obstacles and other robots without exceeding a desired maximum distance from a predefined set of neighbors (flocking constraint). In addition, we show how the proposed algorithm can deal with tracking errors and onboard sensing errors without violating safety and proximity constraints, still providing the conditions for having convergence towards the goal region. To validate the approach, we provide experiments in the field. We tested our algorithm in GNSS-denied environments i.e., a dense forest, where fully autonomous aerial robots swarmed safely to the desired destinations, by relying only on onboard sensors, i.e., without a communication network. This work marks the initial deployment of a fully distributed system where there is no communication between the robots, nor reliance on any global localization system, which at the same time it ensures safety and convergence towards the goal within such complex environments.

Problem

• Distributed swarming without communication or prior environment knowledge
• Ensuring collision-free robot navigation with flocking constraints
• Achieving goal convergence despite sensing errors in GNSS-denied environments

Innovation

• Distributed algorithm without communication
• Onboard sensing and computation only
• Ensures safety and goal convergence