An intelligent companion that helps people with early-stage dementia to always find their way home safely
There are currently over 55 million people living with dementia worldwide, and rising. 
Transport is the primary way to access forms of social participation, cognitive activity, and physical exercise that have been proven to have a delaying effect on early-onset dementia. However, the fear of getting lost is also what prevents people with dementia from engaging in outdoor activities and social events. 
70 % of people with dementia may go missing at least once when walking in public, causing stress and confusion both to themselves and to their caregivers.  
Product Description
Torch is a companion device that always navigates users to their homes. It includes a compass feature with a side light pointing to home and visual, haptic, and sound-based feedback that could be customised to each user's needs. The device evolves with the user's conditions by setting up an accompanying app that allows users to enter their conditions, register their homes, add key landmarks, and customise the types of prompts. This could be done by a family member or the user depending on the level of their familiarity with smartphones. 
The app records the user’s walking patterns, calculates the number of successful journeys, and analyses the mistakes made during their journeys. It also suggests AI-driven alternative options for outputs to improve future journeys. The app user is able to change the outputs following a simple guidance on the app.
Product development process
First, we started with desk research on the relationship between cognitive stimulation and navigation abilities, which we incorporated into our product by linking personal memory triggers and navigation systems. Then we interviewed caregivers and professional medical staff about users' conditions and experiences to find out the major issues around visuospatial abilities. We focused on developing a device that supports the independence of care recipients and differentiated our product from existing products that mainly focus on passive monitoring and are reliant on caregivers.
The process was followed by a simple experiment with users to test different types of visual and audio cues to help them with navigation. We focused on developing simple and optimal cues that would not overload users with too much information. We broke down the cues into visual, audio, and haptic formats with different levels of complexity.
We also talked with experts to get deeper insights, which helped us to focus on integrating the device into users' daily routines by offering various modular options depending on their specific needs and lifestyle preferences to minimise the learning curve.
We conducted a feedback session with a group from Innovations in Dementia. The group consisted of 9 members at the age of 50-70 with early-stage dementia. We got feedback about the device, types of visual cues, and usage cases. Given the progressive and diverse nature of the condition, we incorporated AI into the app that would automatically customise and update prompts informed by analytics and user behaviour, enabling the device to evolve with users.
Exhibition
This project was showcased at the Royal College of Art Dyson building. We got positive feedback from various audiences.

You may also like

Back to Top