Martijn N. Rooker

• Exploration is one of the main research topics in the field of mobile robotics. Multiple robots have been applied for different exploration algorithms In most of the these algorithms the robots exchange data via wireless communication assuming perfect data transmission in the whole environment. Unfortunately, this assumption does not always hold, as e.g. the range of wireless communication is limited. In this thesis Communicative Exploration is presented, which is based on Frontier-Based Exploration. For communicative exploration a group of robots is exploring unknown environments and maintain communication throughout the whole exploration process. The robots form a mobile ad-hoc network that is fully connected for the entire exploration. The motion of the robots is controlled via a heuristic utility function which assigns values to a population of configurations. The configuration with the highest utility value defines the motions of the robots. Different versions of communicative exploration are presented. In the first version a basestation is located outside the environment, from where a human operator can control or supervise the group of robots. During exploration the robots also maintain communication with the basestation. An extension to this version is presented where an improved motion control algorithm is implemented. By calculating a distance transform over the environment the robots are guided better towards the unknown areas of the environment. In a different version the basestation is omitted from the setup to enlarge the reachability of the robots. The group of robots explores the environment acting as a robot pack. Surprisingly, this version introduce situations where the robots get stuck in the environment in deadlock situations. Strategies are introduced that try to avoid or recover from these deadlocks.