Interact at Own Risk! Developing a Prototype of a Robot Hazardous Capability Labeling System

We propose to employ hazard labels to communicate a robot's potentially hazardous capabilities and behaviours to users. Robots can pose a range of ethical and physical safety concerns that users must be aware of when deciding whether to interact. We developed an initial set of key hazards, and a corresponding prototype of a hazard labeling scheme, to demonstrate the potential of this approach. We intend to use this prototype to support exploration of different styles of labeling, and ultimately to develop a more formalized system of robot hazard communication.