Social robot navigation to improve people comfort in collaborative navigation daily tasks
- Ginés Clavero, Jonatan
- Francisco Javier Rodríguez Lera Director/a
- Francisco Martín Rico Codirector/a
Universidad de defensa: Universidad Rey Juan Carlos
Fecha de defensa: 26 de julio de 2022
- Paloma de la Puente Yusty Presidente/a
- José María Cañas Plaza Secretario/a
- Francisco Bellas Bouza Vocal
Tipo: Tesis
Resumen
In recent years, robots have been moving from research laboratories to being part of everyday life in our homes and society. This new reality creates everyday situations where humans interact and collaborate with robots. For such interactions to be satisfactory for humans, they must be natural, comfortable, and safe, mainly if we focus on human-robot interactions with mobile robots. The above conditions require developing robotic systems adaptive to a dynamic environment, such as an environment populated by people and to current social rules. So, the main goal of this thesis by compendium is propose a solution, which integrated into the robot control architecture, allows us to modify the robot behavior according the people's behavior with whom it shares space. There are four fields of study that compose the fundamental part of the proposed solution: Robot Navigation: the robot must navigate in dynamic environments reliably and safely, Human-Robot Interaction: responsible for handling the social systems, such as dialog or proxemics, to collect data from humans, Context Awareness, the robot has been aware of the environment and process it to generate knowledge. Finally, Cognitive Architecture orchestrates the behaviors following principles based on knowledge and/or human-like cognitive processes. The main contributions of the present dissertation are the development of a dynamic solution to represent people and their context in terms of personal space and the development of a novel proxemic zone to improve collaboration tasks between humans and robots, the cooperation zone. These contributions play a part in improving people's comfort during HRI behaviors and have enabled collaborative navigation actions, such as escorting or following a person. Furthermore, the empirical investigations performed in competition scenarios demonstrate the solution's validity and show that it is ready to be deployed in real environments.