Memories through Touch – project update 2

Welcome back to the second in Lyes Oussaiden’s series of ‘Memories’, blogs documenting the journey throughout my research and developments on ‘Memory through Touch’.

Happy new year! The last couple months of 2025 witnessed a significant pivot, including a new approach to tactility with the prototype experience, a new gadget, and a couple new approaches.

I am also now able to commence recruitment. More on that at the end.

Because of the heavy focus on practical development, there’s a lot to talk about. This blog has been split in half, with the next part coming soon. But first, the reasoning for the pivot.

Technical Debt

I previously noted the use of an ‘object tracking’ technology as a means of figuring out where objects are in an environment on a computer. Since the last blog post, I have managed to get this working. However, the raw results were unreliable and choppy.

Video demonstration of a prototype attempting to track a physical bottle in digital space.

While this might appear promising in its prototype state, much of my time would be spent getting stable results instead of co-designing something actually helpful, incurring a technical debt that I would struggle to correct on my own. This route was initially chosen due to a presumed ‘robustness,’ meaning that so long as the video feed is uninterrupted, there would be nothing that you could accidentally break. Unfortunately, it is difficult to get smooth and accurate results. There’s also a risk that a fully built-out system would require the end-user to purchase increasingly expensive hardware to get it working, leading to outcomes becoming inaccessible.

Perhaps this is a route that could be revisited down the line once the technology has advanced or if someone with a strong mathematical background becomes directly involved in this area of research.

Pivoting to RFIDs

I had a chat with a friend, Dr. Naman Merchant, at the end of October to discuss alternative solutions to ‘tracking reminiscence artefacts’ using physical hardware. He spoke to me about various technologies and suggested a concept incorporating RFID tags. Radio-frequency Identification (RFID) is an umbrella term encompassing tech that can be used to identify physical objects. RFID ‘tags’ – which typically come in the form of cards, keychains, and stickers – are used for countless things in everyday life. You might even use them without realising it! They can be found in contactless credit and debit cards, hotel keycards, microchips for your pet, security tags on clothes, biometric passports, car keys, and so on. There’s even a history of them in video games such as with Nintendo’s Amiibo and the Skylanders game franchise. As such, the hardware to make this functional is exceptionally cheap and readily available.

Tags need a way to be read, and I used what we had to hand at the university to prototype a reader.

Using the ‘Arduino’ platform, this gadget captures the identification number of whatever tag is placed onto the reader, gets its state (whether it has been placed or removed), and sends this information through a cable to a connected computer. With a bit of code, it can be used in a similar way to any other kind of game controller.

While this software was functional, it’s largest failing was that it doesn’t represent tactility. I went back to the drawing board, wanting to better understand the problem space.

After workshopping a couple ideas, I was encouraged to develop two prototypes. The first, demonstrating a ‘virtual pet’ experience (building off the avatar showcased in the previous experience). Another, showcasing a ‘slideshow’ experience built on user memories.

After two weeks of tinkering, I technically ended up developing three prototypes, all of which are designed to be interlinked should development on them continue. 

The RFID ‘Platform’

Before developing prototypes, I needed a way for the computer to remember reminiscence objects and share this information across experiences. Drawing inspiration from platforms found in set-top boxes and game consoles, I constructed the “RFID Platform”. The platform can start experiences and process special ‘tag’ stickers. Objects with these stickers can then be read and identified by the reader.

A diagram highlighting the role of the RFID Platform in relation to co-designed experiences.

Functionally, the RFID Platform behaves much like a check-in desk when boarding for a flight. You can ask it where you want to go, you give it the things you need for your reminiscence journey so you can take them with you, and it sends you on your way.

But what if the computer detects a sticker it hasn’t seen before?

The RFID Platform handles this too. I constructed a table of activities, locations, and nouns that can be used to identify objects. While this will not cover every potential object you might want to reminisce over, it allows for a linguistic way to tell the computer what the object is. For instance, differentiating a ‘ball’ for football versus for golf. The RFID Platform uses this table to help you categorise your object.

As these words are curated, they should be safe for use in searching online for resources. Here, they’re used to download pictures and 3D models to allow you to construct a digital version of the reminiscence object accessible within experiences.

One downside of using online databases is that they’re mostly based in the United States, so Scottish colloquialisms have to be localised – meaning that they have to either be translated to their North American equivalent – or have to be handled differently. (For instance, a football in Scotland is different to a football in the US.) However, this is all hidden. As far as you would be able to see, Scottish terms go in, Scottish results come out.

Not every object will be covered by this process. The RFID platform handles this by defaulting back to the previous ‘context’. As I have the ‘location’ and ‘activity’ of every object, previous results can be used and selected by the user if they feel more relevant.

In the absolute worst-case scenario, results are left blank. But with this approach, you still be able to select something even if the connection to the object is more abstract.

A rule I’ve set for the design of experiences is that new objects should be able to be added at any time, and the RFID Platform is the tool that allows this to happen.

What’s next?

Next week, we’ll go into detail about experiences I prototyped for this RFID platform.

If you, a friend, or members of a group you are a part of live with dementia and are interested in contributing to this project, please get in touch at l.oussaiden1900@abertay.ac.uk. You don’t need to know anything about design or technology, just a willingness to share your thoughts and experiences!

Thank you very much for reading. Make sure to check back next week! Keep an ear out and watch this space.

All the best, -Lyes Oussaiden