Detecting Anomalous Behavior of Socially Assistive Robots in Geriatric Care Facilities

Abstract

With the need for geriatric care workers growing faster than can be met, the possibility of socially assistive robots filling this need has garnered increasing attention. This heightened interest in robots as social care workers, however, leads to concerns in detecting possible robot misbehavior. We propose a short questionnaire, based on current elder abuse screening tools, as a method to detect intrusion or misconfiguration in caregiver robots. We focus on misbehavior that can cause psychological or financial harm to the caregiver recipient. We discuss requirements, limitations, and future enhancements.