X-HAZOP: A Family of Techniques for Ethical Hazard Analysis of Assistive Robots

Abstract

Determining the ethical acceptability of assistive robots is a complex task, not least because the ethical hazards of such systems manifest differently across different demographics. Users of these robots are potentially likely to be already marginalized by technology and, thus, vulnerable to a wider range of ethical hazards than other demographics. Adequate assessment and mitigation of the robot’s ethical hazards therefore requires creativity, collaboration, and a range of diverse perspectives. This article presents X-hazard and operability analysis (X-HAZOP), a family, or toolbox, of techniques for conducting ethical hazard analysis of an assistive robot by utilizing structured facilitated workshops. We present the findings from multiple workshops, demonstrating that the use of X-HAZOP techniques with a suitably diverse group of participants improves creation of accessible descriptions of the robot, aids understanding, and leads to an effective identification of a range of ethical hazards.