Innovative Robot Research Aims to Support Dementia Patients and Caregivers

by Nicholas Schou

In a groundbreaking research project at the University of California, Irvine, Dr. Adey Nyamathi, a distinguished professor emerita and founding dean of the Sue & Bill Gross School of Nursing, is leading an initiative that seeks to harness the potential of robotics to improve the lives of elderly patients suffering from dementia. This innovative endeavor, which commenced with a grant from the University of California Office of the President, aims to design a robot capable of recognizing agitation in dementia patients and intervening to prevent falls and reduce caregiver burden.

Dr. Nyamathi, who recently retired after an illustrious 40-year career in the UC system, has been instrumental in assembling a diverse team of experts for this ambitious project. The team includes multiple principal investigators, such as Dr. Amir Rahmani from the School of Nursing and Dr. Nick Dutt from the Bren School of Information and Computer Sciences, as well as Dr. Jung-ah Lee, whose expertise in dementia is critical to the research. In addition, another critical team member is AI systems expert, Dr. Erik Krogh from Pepperdine University as well as a host of computer scientists and engineers from NaviGAIT in Pakistan.

Our goal is to create a robot that can not only recognize when a patient is becoming agitated but also respond in a way that alleviates their distress. This could transform the caregiving experience.”

Dr. Adey Nyamathi

The project, titled “Establishing the Foundations of Emotional Intelligence of Care Companion Robots to Mitigate Agitation Among High-Risk Dementia Patients,” envisions a robot that can measure patient biodata, predict potentially dangerous behaviors, and engage with patients empathetically to avoid negative outcomes. “Our goal is to create a robot that can not only recognize when a patient is becoming agitated but also respond in a way that alleviates their distress,” Dr. Nyamathi explained. “This could transform the caregiving experience.”

Currently, there are approximately 11 million individuals with dementia in the U.S., a figure expected to rise to 14 million by 2060. Dementia patients are four to five times more likely to experience falls, which can lead to severe health complications or even death. “The burden placed on caregivers is immense,” she noted, “and our robot could provide them with much-needed respite, allowing them to focus on their own well-being.”

The initial phase of testing is being conducted in a board and care facility in Santa Ana, California, where the research team has begun deploying cameras to capture audio and visual data on patients’ behaviors and emotional states. “By analyzing this data, we can develop predictive models that alert the robot when a patient may be getting agitated,” Dr. Nyamathi said.

The robot, which resembles a three-foot-tall R2D2 and is equipped with a tablet, and multiple sensors, will be programmed to initiate conversations based on the individual patient’s history and interests. “For example, if the robot senses that a patient named Tommy is becoming agitated, it might say, ‘Tommy, tell me about your friend at Boeing,’ triggering a positive memory that helps distract him from distress,” she explained.

Dr. Amir Rahmani, a co-investigator on the project, highlighted the role of generative AI and large language models (LLMs) in crafting personalized interactions to calm patients. “By leveraging generative AI, we can create dynamic, context-aware conversations that are deeply personal and meaningful to each patient. This storytelling approach not only mitigates agitation but also fosters a sense of comfort and familiarity, making the robot a truly empathetic companion,” Dr. Rahmani said.

“By leveraging generative AI, we can create dynamic, context-aware conversations that are deeply personal and meaningful to each patient. This storytelling approach not only mitigates agitation but also fosters a sense of comfort and familiarity, making the robot a truly empathetic companion,”  Dr. Rahmani said.

The research also involves collaboration with experts in linguistics and AI, including Dr. Sabine Brunswicker from Purdue University, and Dr. Cheonkam Jeong, UCI postdoctoral fellow, who is focused on designing the robot’s conversational models. These interactions will be tailored to each patient’s preferences, allowing the robot to adapt its voice and even display pictures of family members to enhance engagement. “This personalization is crucial,” Dr. Nyamathi emphasized. “The more relatable the robot is, the more effective it will be in providing comfort.”

While the robot has been in development for several years, this project represents a significant leap forward in its application for dementia care. The initial phase of the research is funded by a $300,000 UCOP Noyce grant, which has been extended to a third year to continue exploring the complexities of emotional detection and intervention strategies.

As the team progresses, they aim to validate the robot’s effectiveness in predicting agitation and falls, ultimately paving the way for a practical tool that can improve the quality of life for both patients and caregivers. “This is just the beginning,” Dr. Nyamathi said. “We envision a future where technology and compassion go hand in hand, offering support to patients and caregivers where it’s needed most. With the right tools, we can make a meaningful difference in their lives.”