Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
The widespread deployment of mobile service robots in public environments and the growing use of autonomy as control modality will increase the exposure to hazards for people simply sharing the robot floor. The need to ensure safety becomes an ethical and legal obligation for designers. However, current safety requirements are mostly focused on the end-users of robot services, while pedestrians are often not addressed in safety standard requirements. In this paper, we propose a preliminary method for assessing robot safety with respect to collisions with bystanders. Consisting of three main phases: (1) identification of collision contacts points; (2) estimation of collision injuries; (3) design and control recommendations. We evaluate the robot Qolo - a person carrier robot as an example by simulating collisions with walking pedestrian models, using motion capture and body shape data of a male human adult, and a child. We present results of a simulation comparing multiple post-collision reactions and discuss further work needed for creating a framework on robot safety requirements that address all vulnerable populations.
Pierre Dillenbourg, Elmira Yadollahi, Ana Paiva
,