Enhancing Artificial Intelligence (AI) for Automation and Robotics Through Awareness of Human-Object Interactions

Abstract

Humans interact with objects in complex and diverse ways, through activities such as lifting, picking, handing objects to each other, assembly, taking apart, and repairing. As systems such as robots and intelligent agents become increasingly pervasive in user spaces, artificial intelligence (AI) algorithms driving these systems need to be aware of human interactions with objects to ensure seamless integration for tasks such as human-robot collaboration (HRC) and automation in domains such as manufacturing and healthcare. In this talk, I will discuss our work on studying how humans perceive and interact with objects to create AI algorithms that are aware of human-object relationships. I will talk about how our recent work on creating AI-driven algorithms for automated human-robot collaboration (HRC) tasks from dense multi-modal and multi-person data enhances the state-of-the-art for enabling seamless HRC. Further, I will discuss how our new multi-institutional cross-disciplinary research on "recyclofacturing" transforms manufacturing by enabling metal recyclers to fabricate user products from scrap metal by integrating AI, extended reality, and HRC and workforce training and education toward cyber-enabled manufacturing.

Short Bio

Venue

LCR, O'Relliy