From Crowd Simulation to Robot Navigation in Crowds

This letter presents the result of a study aiming at investigating to what extent the results obtained in the Crowd Simulation domain could be used to control a mobile robot navigating among people. It turns out that Crowd Simulation relies on two assumptions that would not hold for a real mobile robot, a test protocol has therefore been designed in order to thoroughly evaluate how three representative Crowd Simulation techniques would perform when said assumptions are relaxed. The study shows that all those techniques entail safety problems, i.e. they would cause collisions in the real world. The study also highlights the most promising candidate for a transposition on a real mobile robot.