Facoltà di scienze informatiche

Path planning for mobile robots in the real world : handling multiple objectives, hierarchical structures and partial information

Guzzi, Jérôme ; Gambardella, Luca Maria (Dir.) ; Giusti, Alessandro (Codir.)

Thèse de doctorat : Università della Svizzera italiana, 2018 ; 2018INFO010.

Autonomous robots in real-world environments face a number of challenges even to accomplish apparently simple tasks like moving to a given location. We present four realistic scenarios in which robot navigation takes into account partial information, hierarchical structures, and multiple objectives. We start by discussing navigation in indoor environments shared with people, where routes are... Plus

Ajouter à la liste personnelle
    Summary
    Autonomous robots in real-world environments face a number of challenges even to accomplish apparently simple tasks like moving to a given location. We present four realistic scenarios in which robot navigation takes into account partial information, hierarchical structures, and multiple objectives. We start by discussing navigation in indoor environments shared with people, where routes are characterized by effort, risk, and social impact. Next, we improve navigation by computing optimal trajectories and implementing human-friendly local navigation behaviors. Finally, we move to outdoor environments, where robots rely on uncertain traversability estimations and need to account for the risk of getting stuck or having to change route.