Distributed Coverage Control on Surfaces in 3D Space

Adaptive Multi–Robot Coverage of Curved Surfaces

“…It is important to provide an e↵ective deployment and partitioning of space in the first stage, as each robot needs to cover the entire surface of its assigned Voronoi region in the subsequent second stage. The proposed coverage algorithms make use of anisotropic centroidal Voronoi tessellations [8], and extend our previous work [7] with enhanced adaptivity of the Voronoi regions. The Voronoi regions adapt to local anisotropy, which is defined by a tensor field on the curved surface.…”

“…In our work, we consider robots that move directly on the surface [6,7]. Spaces in industrial structures are often narrow and thus forbid robots to work together in close side by side formations.…”

“…Similar to [7,9,10,11], we perform the robot deployment based on a discrete representation and model the environment as a graph. In our previous work [7], we suggested two fundamentally di↵erent approaches for covering a surface, based on discrete geodesic and approximative Euclidean distance computation.…”

“…In our previous work [7], we suggested two fundamentally di↵erent approaches for covering a surface, based on discrete geodesic and approximative Euclidean distance computation. The algorithms in this paper follow up on both approaches in presence of anisotropy.…”

Null space optimization for effective coverage of 3D surfaces using redundant manipulators

“…They also do not address the problem of minimizing costs in the configuration space of the robot and do not consider robotic manipulators. Breitenmoser et al [3] extended 2D coverage for mobile robots to 3D surfaces, by using Voronoi tessellations to map the surface to a 2D plane. Although closely related, they only considered mobile robots moving on the surface and coverage in terms of Euclidean distance.…”