Ali, AbdikadarThaxton, Danielle2021-10-222021-10-222020http://hdl.handle.net/11216/40132020 Celebration of Student Research and Creativity presentation"The last couple of years have seen a strong movement supporting the need of having ethically aligned trustworthy artificial intelligence (AI) within intelligent system-based consumer products, including autonomous cars and robots. This global movement has led to multiple institutional recommendations towards ethical alignment and trustworthy design in AI based autonomous systems and intelligent consumer devices. The IEEE global standards on the design of ethically aligned, trustworthy AI includes a list of guideline principles in this regard. There has been prior research towards finding design flaws and vulnerabilities within various types of consumer robots. However, none of these previous works have studied whether discovering design flaws and security or privacy issues in these robots can help assess their alignment with these IEEE benchmark requirements for trustworthiness and ethical alignment. In attempt to address this gap in existing literature, we have performed a unique experimental study of two educational robots - Zümi and Cozmo. We have found specific vulnerabilities and design weaknesses in these robots within their system functionalities. Our initial research shows that these flaws can lead to hacking, injection attacks, and other robotic malfunctions that might affect the technology users negatively. We conduct a preliminary analysis of how these design flaws can be potentially non- compliant with the IEEE principles for ethically aligned, trustworthy AI. We demonstrate a novel case study of how discovery of design flaws in educational robots can assist in assessing whether their designs align with the new IEEE standards plus principles for robot ethics & trust."en-USAutonomous robotsArtificial intelligenceDesign and technologyInstitute of Electrical and Electronics EngineersPresentation