The first seminar of our Spring 2024 series, featured Dr. Chien-Ming Huang, who gave an engaging, informative, and eye-opening presentation based on a series of recent research on human-robot interaction in healthcare. We also thank Dr. Kathy McDonald for her insightful discussion building on Dr. Huang's work.
Below is the AI-generated summary of the January 12, 2024 seminar, with minor edits made by the AI Workgroup Leadership:
The HBHI-AI workgroup focuses on integrating AI into healthcare workflows. Chien-Ming discussed the evolution and potential of human-robot interaction in healthcare and other industries. He emphasized the significance of designing robot behaviors to facilitate interaction and establishing clear interaction expectations. He also highlighted the challenges and potential issues arising from human-AI interactions.
Integrating AI Into Healthcare Workflows
Chien-Ming discussed the evolution and potential of human-robot interaction in healthcare and other industries. They emphasized the significance of designing robot behaviors to facilitate interaction and establishing clear interaction expectations. They also highlighted the challenges and potential issues arising from human-AI interactions, and shared insights from a study involving radiologists working with an AI system. Finally, Chien-Ming discussed the potential for using technology to engage with children with autism, explaining a setup that uses a robot and a large touch screen to model typical human interaction.
Robot System's Impact on Children's Interaction Skills
Chien-Ming discussed a field deployment of a robot system designed to encourage children's behavior and facilitate interaction skills through daily activities. The robot's presence is intended to improve children's abilities to interact with humans, but Chien-Ming noted that improvements made in the presence of the robot did not persist after the system was removed. This raised questions about the optimal duration of exposure to the system in order to see sustained benefits. The meeting also experienced connectivity issues, with Tinglong attempting to troubleshoot the problem.
Robot-Assisted Behavior and Exercise for Autistic Children and Older Adults
Chien-Ming shared the findings from two studies. The first focused on the use of robots to assist children with autism. It was found that the benefits of robot-facilitated behavior or exercise were not sustained after a 30-day deployment. This prompted questions about the optimal duration for deploying assistance to ensure sustained benefits. The second study explored the use of robots for older adults. The research team took a human-centered approach, focusing on understanding the real struggles of older adults and designing robotic systems to improve physical activity engagement. They found that older adults have various motivations for physical activity, but also face barriers such as embarrassment and the inability to keep up with group sessions. The team is now designing and deploying prototypes based on these insights.
Human-AI Interaction and Emotions
Tinglong shared their background in robotics and their interest in human-AI interaction. Kathy recounted their experience with a chatbot from United Airlines and expressed curiosity about how human behavior changes when interacting with non-human technologies. They observed that their frustration level increased faster than usual during these interactions. Zhiqing mentioned a conversation with a Microsoft representative about their work on emotional aspects of AI, including how different chatbot models respond to emotions. Kathy concluded by sharing how their behavior changed when interacting with the chatbot, particularly when they were frustrated, and wondered about the potential for empathetic and decision-making responses in chatbots.
Technology and Robotics in Healthcare Discussion
Kathy expressed interest in the robot's problem-solving process and its technical aspects. Chien-Ming discussed the process of repurposing existing technology for older adults and the challenges they faced in understanding how people used technology. They also mentioned a social prototype robot and a new project involving an Exoskeleton to encourage physical activity. They further discussed the integration of a large language model with voice assistants to enhance interaction and shared findings from a lab study on self-diagnosis.
Voice Interaction Technology in Healthcare and Manufacturing
Chien-Ming discussed the application of voice interaction technology in healthcare, particularly for older adults. They emphasized the importance of monitoring symptoms and suggested using smart speakers and recorders for this purpose. Chien-Ming demonstrated the technology's ability to handle natural speech inputs and proposed linking it to robots for potential future developments. They also touched upon the possibility of using this technology to increase work satisfaction through human-robot collaboration in manufacturing and assembly. Chien-Ming highlighted the challenge of silence in human-robot interactions and proposed the idea of getting robots to engage in casual conversation with humans to make work more enjoyable and potentially increase productivity.
Real-World Challenges in Deploying Healthcare Robotics
The discussion then moved to the challenges of implementing robotics in diverse home environments due to the complexity of different preferences and home layouts. Chien-Ming explained that their project was supported by the NSF and is still in the early stages of technology development. Tinglong inquired about the manpower required for the maintenance and operation of these robotic systems, which Chien-Ming acknowledged would require effort.
