**3.2 Path planning**

The multi-robot CPP approach is being pursued in various studies describing challenging problems and proposed solutions. All of these approaches have a similar goal of providing a collision-free path that achieves the full extent of a structure or area. As a representative method, the visibility graph is used for path generation [17]. Visibility graphs contain a set of points and obstacles where nodes represent locations, and edges consist of line segments that do not pass-through obstacles. Examples of visibility graphs and generated paths are shown in **Figure 5**.

Geometric approaches such as shortest Euclidean path finding and polygonal area coverage are used in many domains. The most popular geometric-based method in multi-robot CPP is the Voronoi diagram. In [19], a dynamic path planning approach was proposed for heterogeneous multi-robot sensor-based coverage (DPP MRSBC)

**Figure 5.** *Examples of visibility graphs and generated paths [18].*

considering energy capacity. The environment was modeled as a generalized Voronoi diagram (GVD) that obscures the edges of the diagram. The proposed algorithm starts with an undirected graph and creates a directed subgraph.
