The number of people living with dementia is expected to rise significantly in the coming years, potentially doubling in the United States by 2050, according to projections. One of the challenges faced by individuals with dementia is the loss of long-term episodic memories, including important details such as first-day-of-school experiences or sensory information like sights and smells.

Episodic memories are stored in the hippocampus, which is often affected by dementia, causing difficulties in recalling recent events. With no known cure for dementia, treatment focuses on slowing its progression and enhancing quality of life.

To address these challenges, engineers at the University of Waterloo have developed a robot designed to assist people with dementia in locating misplaced items such as medicines and glasses. Dr. Ali Ayub, a postdoctoral fellow in electrical and computer engineering, and his team presented their research at the ACM/IEEE International Conference on Human-Robot Interaction this year.

Dr. Ayub explained, “The long-term impact of this is really exciting. Users can benefit not only from a companion robot but also from a personalized companion robot that promotes independence.” While the initial focus is on individuals with dementia, the team envisions the technology being applicable to anyone who needs help tracking misplaced items in the future.

The researchers began their project with a Fetch mobile manipulator robot, equipped with an arm for object manipulation, multiple sensors, and cameras. They programmed the robot with an object-detection algorithm that enables it to identify and keep track of specific items within its field of vision.

Users can select the objects they want the robot to track through a user-friendly graphical interface on their smartphone or computer. When an item needs to be found, the user can initiate a search via the interface, prompting the robot to indicate where and when the object was last seen.

While the technology may appear complex, Dr. Ayub assures that it is user-friendly and accurate. The team plans to conduct real-world testing, initially with individuals without disabilities and later with dementia patients.

The study detailing the research was published in the ACM journal. The researchers hope that their episodic memory system, integrated with a mobile manipulator robot, will not only track and store various household objects but also provide a visual representation of these items. Encouraging results from system evaluations have prompted the team to continue investigating the system through co-design and user studies in the future.

By Impact Lab