Edinburgh Tram POS
A self initiated concept project that tackles the flawed experience of buying a tram ticket in Edinburgh.
— —
Role: Designer.
Existing Design
Concept Design Proposal
I live in Sweden and visit family in the U.K. often.
The trip is long and involves 6 different modes of transport. It sounds worse than it really is. Bring a book, snacks and you’re good.
But there is a step in the journey I can’t ignore any longer.
The Edinburgh Tram Ticket Machine.
Context
The opening screen isn’t terrible. It’s not going to win any design awards but it does the job.
The pain, so much pain.
But the station selection screen hurts my head.
Here’s the thing
The people who use these machines the most, are often the least familiar with the city.
Locals/commuters tend to have monthly passes.
So for tourists & rookies, this screen is a spaghetti of names that aren’t relatable.
Not just that but there’s so many choices to pick from.
Simply put:
Too busy and unfamiliar.
Early Sketches
My first priority was to simplify the screen.
I wanted to show all the stations in a less intense way.
Early sketches of possible layout
These concepts could have been fun…
But I’m sure all of us have experienced a kiosk screen before.
They’re often slow and glitchy.
Adding a big scrolling interaction would likely have been an even more frustrating experience.
Was there a better way to reduce the options?
Every station has a machine on both sides of the track.
Which means we can suggest a handful of stations the customer will likely want to go to from the station they’re at.
And these smart suggestions would vary at each station.
i.e. We’re not going to show Edinburgh Airport as a popular station if you’re travelling in the opposite direction.
Being smarter with context
Sketches focused on “popular stations”
I had a pretty good idea of the direction I wanted to take things but I wanted to incorporate AI to get an alternative perspective.
I experimented with a handful of AI platforms, including Lovable, Magic Patterns and Stitch.
The results were interesting and it also raised the idea of having popular stations, however the output was average. I imagine with more tokens and time, I could have refined them to be much better.
Involving AI
Ai Design Concepts
This project was an experiment with a cap limit of 1,500 jackets.
Would it work? Did customers truly want this?
Our site crashed on launch.
The demand was enormous. It completely exceeded our expectations and our project was repeatedly highlighted and applauded internally by senior management.
We did an incredible job with limited resources. In hindsight, I’d have allocated more time to early user testing — it could have led us to a more elegant solution, faster.
impact & Reflections
More projects
Edinburgh Trams Ticket Machine
Hospital AR Glasses