New end-to-end task agnostic AI systems have transformed the ability of wearable robotics to provide meaningful assistance to users across a diverse range of daily activities. This talk will present best practices for structuring real-time deep learning systems for wearable robotics that integrate seamlessly with the user’s underlying physiology. Outcome measures from related human subject experiments with test scenarios of these systems will be presented. Building on this, the emerging science of transfer learning has transformed our ability to deploy large-scale deep learning models at scale across diverse sets of wearable robots, tasks, and populations, which removes the need for new research teams and companies to collect numerous amounts of expensive, labeled data to train these systems. Our groups latest advances towards leveraging transfer learning to create foundational human motion models will be discussed. Lastly, these techniques can be applied to a very broad set of practical problems, and three unique deployment scenarios will be discussed. First, our research team recently completed a clinical trial in mobility impaired individuals from stroke using hip exosuits to enhance daily activities of living. Second, strenuous industrial tasking can lead to injuries which our team has found indications that these can be mitigated through effective clothing-integrated robotic systems that provide task agnostic assistance. Lastly, balance and stability are a primary struggle for many individuals with mobility deficits yet are very hard problems for wearable robotics to address. I will highlight some of our team’s recent successes by leveraging predictive internal state estimation in combination with wearable exoskeletons to improve balance during unstable gait.

Aaron Young is an Associate Professor and Woodruff Faculty Fellow in the Woodruff School of Mechanical Engineering at Georgia Tech and has directed the Exoskeleton and Prosthetic Intelligent Controls (EPIC) lab since 2016. Dr. Young received his MS and PhD degrees in Biomedical Engineering with a focus on neural and rehabilitation engineering from Northwestern University in 2011 and 2014 respectively. He received a BS degree in Biomedical Engineering from Purdue University in 2009. He also completed a post-doctoral fellowship at the University of Michigan in the Human Neuromechanics Lab working with lower limb exoskeletons and powered orthoses to augment human performance. His research area is in advanced control systems for robotic prosthetic and exoskeleton systems for humans with movement impairment. He combines machine learning, robotics, human biomechanics, and control systems to design wearable robots to improve the community mobility of individuals with walking disability. He has received an NIH New Innovator Award (DP2), NIH NCMRR New Investigator award and IEEE New Faces of Engineering award, and his EPIC lab group won the International VIP Consortium Innovation Competition. He is a Senior Member of the IEEE and serves as an Editor / Associate Editor for BioRob and Rehab Week/ICORR.

Given the explosion of autonomy and artificial intelligence (AI) across transportation and defense systems, medicine, and business settings, the need for humans as supervisors of and collaborators with this often-opaque technology is also rapidly growing. Autonomous systems are often brittle and need humans for their more abstract abilities in knowledge synthesis and judgment, requiring coordination and teamwork, often in unexpected ways. This talk will focus on function allocation between humans and AI and how we should think about designing systems in the future that leverage the strengths of both to develop more robust systems.

Professor Mary (Missy) Cummings received her B.S. in Mathematics from the US Naval Academy in 1988, her M.S. in Space Systems Engineering from the Naval Postgraduate School in 1994, and her Ph.D. in Systems Engineering from the University of Virginia in 2004. A naval officer and military pilot from 1988-1999, she was one of the U.S. Navy's first female fighter pilots. She is a Professor in the George Mason University College of Engineering and Computing and directs the Mason Responsible AI program as well as the Mason Autonomy and Robotics Center (MARC). She is an American Institute of Aeronautics and Astronautics fellow and a member of the Virginia Academy for Science, Engineering and Mathematics. Her research interests include the application of artificial intelligence in safety-critical systems, assured autonomy, human-systems engineering, and the ethical and social impact of technology.

To prevent social accidents, autonomous vehicles need to respond implicitly to other driver and pedestrian behaviors as human drivers would. To make such contingent interaction possible, we have instrumented and collected empirical interaction behavior between human drivers and pedestrians in simulated conditions to model interactive responses AVs should exhibit in response to the behaviors of other road users. So far, our human participant studies using AVs using contingent interaction models outperform models which consistently yield or don’t yield. This suggests that AVs should incorporate socially familiar driving patterns through contextually-adaptive algorithms to improve the chances of successful deployment and acceptance in mixed human-AV traffic environments.

Wendy Ju is an Associate Professor of Information Science and an inaugural faculty member of Cornell's new campus-wide multidisciplinary Design Tech department at Cornell University. She researches and designs human interaction with automated systems. Professor Ju has innovated numerous methods for early-stage prototyping of automated systems to understand how people will respond to systems before the systems are built. She has a PhD in Mechanical Engineering from Stanford, and a Master’s in Media Arts and Sciences from MIT. Her monograph on The Design of Implicit Interactions was published in 2015.