DMVC-Tracker: Distributed Multi-Agent Trajectory Planning for Target Tracking Using Dynamic Buffered Voronoi and Inter-Visibility Cells

TL;DR

This work tackles multi-agent aerial target tracking in cluttered environments by enabling distributed planning with time-varying Dynamic Buffered Voronoi Cells (DBVC) and Dynamic Inter-Visibility Cells (DIVC). It couples these cells with Bernstein polynomial motion primitives and a sample-check-select pipeline to produce collision- and occlusion-free trajectories in real time, including 3D extensions. The main contributions are the DBVC/DIVC formulations, a less conservative feasibility check, and a fast trajectory generator that achieves real-time performance on commodity hardware, demonstrated in simulations and hardware experiments with dozens of obstacles. The results show higher success rates and reliable target visibility compared to state-of-the-art baselines, highlighting the practical potential for scalable distributed tracking in dynamic environments.

Abstract

This letter presents a distributed trajectory planning method for multi-agent aerial tracking. The proposed method uses a Dynamic Buffered Voronoi Cell (DBVC) and a Dynamic Inter-Visibility Cell (DIVC) to formulate the distributed trajectory generation. Specifically, the DBVC and the DIVC are time-variant spaces that prevent mutual collisions and occlusions among agents, while enabling them to maintain suitable distances from the moving target. We combine the DBVC and the DIVC with an efficient Bernstein polynomial motion primitive-based tracking generation method, which has been refined into a less conservative approach than in our previous work. The proposed algorithm can compute each agent's trajectory within several milliseconds on an Intel i7 desktop. We validate the tracking performance in challenging scenarios, including environments with dozens of obstacles.

Figures (6)

• Figure 1: Hardware demonstration of multi-agent target tracking.
• Figure 2: DIVC fomulation
• Figure 3: Comparison of the feasibility checks. Yellow: an area where the agent either cannot see the center of a target or collides with the target or an obstacle. Green: an area where bpmp-tracker considers the collision- and occlusion-free area. Blue: an expanded feasible area by the proposed method.
• Figure 4: Left: Three trackers (blue: ours, orange: fei-gao-wild) follow the red target among black static obstacles. Right: Three blue trackers follow the red target among forty green dynamic obstacles. Magenta regions represent DIVCs.
• Figure 5: Target tracking experiments. The total flight paths of Crazyflies serving as trackers (blue), target (red), and dynamic obstacles (green), along with boxes (black), are plotted in a top-down view. The purple areas are the accumulated histories of the Lines-of-sight connecting the target and the trackers.

Theorems & Definitions (2)

• proof
• proof