Human Autonomy as a Design Principle for Socially Assistive Robots
This work tackles the problem of preserving human autonomy in human-robot interaction for vulnerable users, representing an incremental step in design principles.
The paper addresses the risk that increasing robot autonomy may reduce human autonomy, especially for vulnerable populations like older adults, and proposes centering human autonomy in socially assistive robot design by defining autonomy and providing architectural requirements, exemplified by the Assist architecture.
High levels of robot autonomy are a common goal, but there is a significant risk that the greater the autonomy of the robot the lesser the autonomy of the human working with the robot. For vulnerable populations like older adults who already have a diminished level of autonomy, this is an even greater concern. We propose that human autonomy needs to be at the center of the design for socially assistive robots. Towards this goal, we define autonomy and then provide architectural requirements for social robots to support the user's autonomy. As an example of a design effort, we describe some of the features of our Assist architecture.