A research prototype with a hard ceiling
Reminiscence therapy — revisiting personal memories through photographs, music, and familiar places — is a long-established practice in dementia care. At Ontario Tech University, a research team led by Dr. Winnie Sun of the Faculty of Health Sciences, with virtual-reality researcher Alvaro Quevedo of the Faculty of Business and Information Technology, had built something genuinely promising around it: a VR environment that places a person living with dementia inside a familiar room, surrounded by their own past.
The research was real and peer-reviewed — published in IEEE Xplore and presented at the Serious Games and Health Applications conference — and Ontario Tech has shared the early results publicly. A pilot run with the Alzheimer Society of Durham Region had shown people with dementia responding warmly: returning to the experience eagerly, and sometimes recalling memories that had been out of reach for years.
But the prototype had a limitation that capped everything around it.
Every memory in that virtual room — every photo on the wall, every song, every video — had to be written into the application by a developer before it was built. The experience was personal in principle and fixed in practice. To bring a different person’s world into the room, someone had to rebuild the app. A recreation therapist couldn’t add a photograph; a daughter couldn’t add her father’s favourite song. The thing that makes reminiscence therapy work — that it is personal — was the one thing the software couldn’t do on its own. Making content dynamic rather than hard-coded is a genuinely hard engineering problem: the kind that outlives the timelines and rotating teams a research grant is built to support.
What we built
Ontario Tech brought us in as the project’s development partner to close that gap.
Make the content something anyone can change. We built a no-code content pipeline — closer to a shared drive than a piece of software. A family member or a member of the care team drops personal media into a profile: photographs, music, video, panoramic images. The system sorts each file by type and places it where it belongs inside the virtual room — images into frames and albums, music into a player, video onto a virtual screen. No developer, no rebuild. The room becomes that person’s room, and it keeps changing as their family adds to it.
Design the interaction for the actual user. The experience runs on a standard VR headset and uses hand tracking — no controllers to learn. But a single way in is a single point of failure: when hand tracking didn’t work for a participant, they had no way into the experience at all. So we added a second method — a “hand-as-controller” input, where simple controls appear on the user’s open palm and a touch of finger to palm registers as a press, with real, physical feedback. For someone with limited dexterity or range of motion, it turns an unfamiliar gesture into a familiar one. And for users who can’t comfortably wear a headset, a companion non-immersive version delivers the same experience on a screen.
Rebuild the foundation so the next team can build on it. The prototype had grown by folding separate systems into one project — a structure that was the source of much of its fragility. We separated them into clean components that communicate with each other, and where a proven foundation already existed for the VR delivery layer, we built on it rather than writing one from scratch. The result is a codebase the next student or developer can actually pick up.
The outcome
The rebuilt tool went through usability testing with the people who matter most: individuals living with dementia, their family caregivers, and the healthcare providers who would use it alongside them. The response was clear and encouraging — people with dementia found the virtual environment intuitive and engaging rather than confusing, which is never a given with new technology and this population.
The research continues, led by Dr. Sun’s team in partnership with the Alzheimer Society of Durham Region and, through Ontario Tech’s Advancement for Dementia Care Centre, the Ontario Shores Centre for Mental Health Sciences. The next phase — gamified, activity-based experiences, richer environments, a conversational guide — is being built on a foundation designed to be extended rather than replaced.
What we delivered was not a new idea. The idea, the clinical insight, and the evidence belong to Dr. Sun’s research team. What it needed was the engineering to make it real: a tool a family can fill with a life, that a care team can use without a developer in the room, and that the next group can keep improving. That stretch — from a prototype that proves a concept to a product that survives contact with the world — is the work Peppermint Labs exists to do. Peppermint Labs is a Toronto-based digital health and research commercialization consultancy, working with university research teams and clinical partners across Canada.