Multi-UAV Path Following using Vector-Field Guidance

A team of researchers has published a new paper, "Multi-UAV Path Following using Vector-Field Guidance," proposing a fully decentralized framework for coordinating swarms of uncrewed aerial vehicles (UAVs). The system, developed by Gautam Kumar, Amit Shivam, and Ashwini Ratnoo, tackles the core challenges of multi-drone operations: getting each vehicle to follow a predefined path while avoiding collisions and maintaining consistent, uniform spacing between all agents. Unlike centralized systems that rely on a single control point, this decentralized approach allows each drone to operate autonomously based on local information, making the swarm more robust and scalable.

The proposed method combines three key components. First, a vector-field guidance law acts like an invisible force field, pulling each UAV toward the desired reference flight path. Second, a novel "rotational repulsion" mechanism enables collision avoidance; drones sense their neighbors' relative distance and bearing and adjust their trajectory to steer clear, even as they converge on the same path. Finally, an inter-UAV spacing error-based velocity control law dynamically adjusts each drone's speed to achieve and maintain perfect uniform separation along the path. The researchers provide formal mathematical proofs, establishing analytical guarantees that collisions will be avoided and spacing errors will converge to zero.

The efficacy of this coordinated guidance system was validated through extensive numerical simulations. This research, submitted to the 2026 Modeling, Estimation and Control Conference (MECC), represents a significant step in multi-agent systems. By solving the dual problems of safety (collision avoidance) and order (uniform spacing) in a decentralized manner, it paves the way for more reliable and complex autonomous drone applications, from precision agriculture and infrastructure inspection to advanced delivery logistics and aerial light shows.