Bachelor's and Master's Thesis Supervision

The goal is to develop a simulator for patrolling games, where a mobile patroller moves among protected targets represented as vertices in a directed graph. Patroller's moves are controlled by a precomputed randomized strategy. The goal is to visualize the strategy, allow for dynamic changes in the structure of the graph and adapt the strategy on the fly.

Contact: Antonín Kučera, Vít Musil, Vojtěch Řehák

In robotic patrolling, an autonomous device (such as a drone) strives to detect undesirable activities in a predefined set of targets. These activities can be initiated either by an adversarial attacker or spontaneously by nature, which is formally reflected in the objective function used for evaluating the quality of the patrolling strategy. A challenging open problem is how much the chosen type of environment (adversarial or non-adversarial) influences the design of the patroller's moving strategy.

The main goal of the thesis is to answer this question by providing rigorous (theoretical or experimental) evidence.

Contact: Antonín Kučera