This work proposes a contact management approach using Pepper robot, which focuses on safety constraints for navigation through cluttered environments. Firstly, we conduct an analytical study to identify the most common undesired physical contacts between humans in crowded scenes. Based on that, a set of recommendations for robot reaction is provided. Special emphasis is given to contact detection and reaction by proposing a sensorless detection method and different body compliance strategies, respectively, that match the safety guidelines proposed. Finally, an experimental validation is conducted in a controlled environment and through user studies.
CITATION STYLE
Garcia, F., Mazel, A., & Cruz Maya, A. (2019). Safe Human-Robot Interaction Through Crowd Contact Video Analysis. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11876 LNAI, pp. 588–598). Springer. https://doi.org/10.1007/978-3-030-35888-4_55
Mendeley helps you to discover research relevant for your work.