Multi-agent Aerial Monitoring of Moving Convoys using Elliptical Orbits

Aseem Vivek Borkar, Girish ChowdharyICRA 2021 · DOI

Introduction

Monitoring a moving ground convoy from the air is a fundamental challenge in surveillance and security robotics. Traditional approaches often rely on circular orbits around a single target point. However, convoys are typically elongated, making circles inefficient; they either fail to cover the ends of the convoy or waste significant flight time over empty space. This paper presents a cooperative control scheme that enables a team of Unmanned Aerial Vehicles (UAVs) to monitor a moving convoy by maintaining a perfectly spaced, time-varying elliptical formation. This approach optimizes coverage and enhances tracking performance, particularly during sharp or unpredictable convoy maneuvers.

Methodology

The proposed system integrates three core components: dynamic orbit generation, enhanced guidance, and decentralized coordination.

Dynamic Orbit Computation

To adapt to the convoy's movement, the system continuously calculates a minimum-area "bounding box" around the ground vehicles. An elliptical orbit is then fitted to this box. A critical feature of this algorithm is the enforcement of "feasibility" constraints: the resulting ellipse's curvature is limited to ensure that the drones, constrained by their minimum turn radii, can accurately follow the path.

Enhanced Vector-Field Guidance

A primary contribution of this work is the improvement of vector-field guidance laws for non-circular paths. Standard constant-gain guidance laws often struggle with the high-curvature regions at the ends of an ellipse, leading to tracking "overshoot." The authors introduced a curvature-sensitive term into the guidance law. This adjustment makes the drones more responsive to the sharp turns of the ellipse, ensuring tighter adherence to the desired orbit.

Comparison of vector fields for constant gain vs. proposed guidance law.

Parametric Coordination Strategy

Maintaining a uniform formation around a shifting, stretching ellipse is complex. Instead of coordinating based on physical arc length, the drones use the ellipse's parametric angle. By designating one drone as a reference leader, the others adjust their linear speeds within physical limits to eliminate parametric separation errors. This decentralized approach ensures the team eventually achieves an equi-spaced distribution around the moving target.

Experimental Validation

The effectiveness of the proposed control scheme was validated through a series of simulations and real-world experiments. Initial MATLAB simulations demonstrated that the curvature-sensitive guidance law significantly reduced tracking errors compared to traditional methods. High-fidelity Software-In-The-Loop (SITL) simulations using ROS-Gazebo further confirmed the algorithm's compatibility with complex 6-DOF flight dynamics.

Hardware experiment setup and resulting trajectories for multi-agent convoy monitoring.

Final validation was conducted in a motion capture arena using a team of Bitcraze Crazyflie quadrotors monitoring a convoy of custom-built ground robots. In these tests, the aerial team successfully converged to a stable, equi-spaced formation in approximately 61 seconds, even as the convoy navigated a series of waypoints. The controllers effectively managed the drones' linear speeds (between 0.3 and 0.7 m/s) and angular velocities, satisfying all physical constraints.

Conclusion

This research demonstrates that elliptical orbits provide a more economical and effective means of monitoring elongated moving targets than standard circular formations. By incorporating local curvature into the guidance law and utilizing a parametric coordination strategy, the system achieves robust tracking and formation stability. Future work will focus on addressing communication constraints, such as limited range and signal loss, and developing reconfigurable formations to allow for the dynamic addition or removal of agents during a mission.

Summary updated

BibTeX
@inproceedings{borkar2021multi,
  author = {Aseem Vivek Borkar and Girish Chowdhary},
  title = {{Multi-agent Aerial Monitoring of Moving Convoys using Elliptical Orbits}},
  booktitle = {IEEE International Conference on Robotics and Automation},
  year = {2021},
  doi = {10.1109/icra48506.2021.9561492},
}
Figure