Theoretical and Experimental Collaborative Area Coverage Schemes Using Mobile Agents

Sotiris Papatheodorou, Anthony TzesIntechOpen 2019 · DOI · OpenAlex

Abstract

This research focuses on the development of collaborative control schemes for mobile ground robots for area coverage purposes. The simplest scheme assumes point omnidirectional robots with heterogeneous circular sensing patterns. Using information from their spatial neighbors, each robot (agent) computes its cell relying on the power diagram partitioning. If there is uncertainty in inferring the locations of these robots, the Additively Weighted Guaranteed Voronoi scheme is employed resulting in a rather conservative performance. The aforementioned schemes are enhanced by using a Voronoi-free coverage scheme that relies on the knowledge of any arbitrary sensing pattern employed by the agents. Experimental results are offered to highlight the efficiency of the suggested control laws.

Introduction

The "area coverage" problem is a fundamental challenge in robotics: how should a team of mobile sensors position themselves to monitor a region as effectively as possible? While classic solutions often assume ideal conditions—identical circular sensors and perfect localization—real-world applications involve heterogeneous robots, noisy localization, and non-circular sensor footprints (such as cameras or LIDAR). This paper provides a theoretical framework and experimental validation for these constraints using distributed control, where each robot makes decisions based only on its immediate neighbors.

Distributed Control Schemes

The authors propose three distinct schemes to address increasing levels of model complexity:

  1. Heterogeneous Omnidirectional Sensing: For robots with different circular sensing radii, standard Voronoi partitions are replaced with Power Diagrams. These diagrams account for the "weight" (sensing range) of each robot, ensuring that agents with superior sensors take responsibility for larger areas.
  2. Positioning Uncertainty: To handle localization errors, the Additively Weighted Guaranteed Voronoi (AWGV) diagram is employed. This approach defines a "guaranteed sensed region" based on maximum possible positioning error, prioritizing certainty over total area.
  3. Anisotropic (Non-Circular) Sensing: For arbitrary sensor shapes, the authors move away from standard tessellations toward a tessellation-free partitioning. Points seen by multiple agents are relegated to a "common region," while agents focus on points only they can observe.
Geometric partitioning of the sensing region for heterogeneous agents, illustrating the boundaries between cells and the sensing region.

For all three schemes, gradient-based control laws were derived to guarantee a monotonic increase in the coverage objective over time.

Simulation and Experimental Results

The framework was validated through both MATLAB simulations and physical experiments using three AmigoBots. Localization in the lab was provided by an external camera using ArUco fiducial markers.

Key Insights

  • The Circle Approximation Trap: Approximating complex elliptical sensors as simple circles led to a 62.4% decrease in effective coverage in simulations, highlighting the necessity of the anisotropic model.
  • Efficiency: In simulations with eight heterogeneous agents, the system achieved over 90% sensor utilization ($H^a$) and nearly 89% total area coverage ($H^r$).
  • Experimental Success: In physical trials, sensor utilization increased from 83.70% to 98.95%, successfully handling positioning errors of up to 0.032 meters using the AWGV framework.
Simulation results for an eight-agent system showing (a) initial state, (b) trajectories, (c) final coverage configuration, and (d) the temporal convergence of the coverage metric H.

Conclusion

This research demonstrates that distributed control laws can effectively manage the complexities of heterogeneous robot teams. While the current proofs rely on convex regions and static environments, the successful experimental implementation proves the robustness of the AWGV and Power Diagram approaches. Future work will focus on non-convex environments, complex vehicle dynamics, and dynamic space density functions.

Summary updated

Abstract

This chapter is concerned with the development of collaborative control schemes for mobile ground robots for area coverage purposes. The simplest scheme assumes point omnidirectional robots with heterogeneous circular sensing patterns. Using information from their spatial neighbors, each robot (agent) computes its cell relying on the power diagram partitioning. If there is uncertainty in inferring the locations of these robots, the Additively Weighted Guaranteed Voronoi scheme is employed resulting in a rather conservative performance. The aforementioned schemes are enhanced by using a Voronoi-free coverage scheme that relies on the knowledge of any arbitrary sensing pattern employed by the agents. Experimental results are offered to highlight the efficiency of the suggested control laws.
BibTeX
@inproceedings{papatheodorou2019theoretical,
  author = {Sotiris Papatheodorou and Anthony Tzes},
  title = {{Theoretical and Experimental Collaborative Area Coverage Schemes Using Mobile Agents}},
  booktitle = {IntechOpen},
  year = {2019},
  doi = {10.5772/intechopen.78940},
}
Figure