In the last decade, edge computing emerged as a paradigm allowing close-to-the-source processing. It brings multiple benefits for mission-critical applications especially that they cannot tolerate delays, downtime or failure. Particularly, in post-disaster management, where the network is scarce and access to the Internet might not be possible, edge servers can be embarked on the robots that are exploring the disaster area transforming them to edge-enhanced devices. When required, they run resource-intensive tasks and provide local decisionmaking to other constrained devices in the disaster area such as a Wireless Sensor Network (WSN). The robots are batterypowered and are used for a mission-critical application where sensitivity and low tolerance for delays are crucial. Therefore, these edge-enhanced devices must be properly managed to meet the needs of the application. In this paper, a Fuzzy Inference System (FIS)-based approach is proposed enabling robots to decide, whenever a resource-intensive task must be executed, whether they can stop exploring the disaster area and act as edge servers. Six parameters reflecting the status of the network and the application are used by the FIS for the decisionmaking. Preliminary simulation results show that, in the proposed approach, the energy consumption in the network is 1.7 times less than the baseline network.