Exploration by Flooding Area Coverage with Settling Agents
Work towards PhD degree under the supervision of Prof. Joseph Ben-Asher (AE/Technion) & Prof. Alfred Bruckstein (CS/Technion)
Department of Aerospace Engineering
Technion – Israel Institute of Technology
Rescue workers coming to a rescue scene usually do not have an up-to-date map of the region nor a way of navigating inside it. Solutions, be they based on swarms or not, must address the following three problems. First, movement of robots may be hampered by rubble on the ground or other unexpected obstacles. Next, lack of a common reference frame makes coordination based on common landmarks difficult or even impossible. Finally, in such collapsed or unknown rescue regions, severe multi-path and signal attenuation limits radio frequency communication between the agents. We consider several algorithms for exploring and filling an unknown, connected region, by simple, airborne agents. The algorithms are designed with an indoor, (partially) collapsed and unknown rescue scene in mind The agents are assumed to be identical, autonomous, anonymous and to have a finite amount of memory. The region is modeled as a connected sub-set of a regular grid composed of square cells. The algorithms described herein are suited for Micro Air Vehicles (MAV) as these air vehicles enjoy an unobstructed view of the ground below and can move above the ground. The agents explore the region by applying various action-rules based on locally acquired information and some of them may settle in unoccupied cells as the exploration progresses. Settled agents become virtual pheromone traces for the exploration and coverage process, beacons that subsequently aid the remaining, and still exploring, mobile agents. We also introduce a backward propagating information diffusion process as a way to implement a deterministic indicator of process termination. Complete covering of the graph in finite time is guaranteed when the size of the region is fixed and there are no faults. Bounds on the coverage times are derived. Extensive simulation results exhibit good correspondence to the theoretical predictions.
The talk will be given in English
Mon, 04-07-2022, 13:30-14:30 (Gathering at 13:15)Classroom 165, ground floor, Library, Aerospace Eng. & https://technion.zoom.us/j/2770163012
Light refreshments will be served before the lecture