Cooperative UAV-UGV modeled by Petri Net Plans specification

Aerial and ground vehicles working in corporation are crucial assets for several real world applications, ranging from search and rescue to logistics. Here we turn our attention to high level languages or team plans [3, 2, 4, 1] for modeling a cooperative scenario, where the UAV/UGV team should operate in tight cooperation to perform a joint task. In this context, by tight coordination we mean that robots must continuously synchronize their individual actions to successfully perform the joint task. This is because the joint task imposes execution constraints to a vehicles that might depend on the state of the other vehicle. For example, in our landing scenario, the UAV must know the intended future locations of the UGV to properly plan a trajectory so to smoothly land on the UGV. In more detail we investigate the use of the Petri Net Plans (PNPs) specification [4] to specify the collaborative landing task. There are several benefits related to the use of the PNPs framework: first it provides a rich graphical representation that helps the designers to create plans with minimal effort, second the generated plans can be monitored during the execution, third PNPs support well-defined structures for handling tight coordination and on-line synchronization in multi robot systems. In summary, our main contributions to the state of the art are as follow: