Stuttgart's current ticket machines use an interface based on physical buttons that is complicated to understand and inaccessible for many.
Research driven, I created an interface for a potential future ticket machine based on a screen to allow people like the over 8 million visitors of the city to buy tickets quickly and transparently.
Observing customers buying tickets on ticket machines gave me the opportunity to understand the target audience, which are mostly travelers and non-regular public transport users. Additionally I found video interviews online of people using the machines, helping me creating personas.
Based on this research I created an as-is-scenario. The important take-aways were problems with users taking a long time to find the correct input for single ticket destinations, understand what the icons on the buttons mean, as well as understand for how many zones they need daily tickets.
The display is small, hard to read, and on the top of the machines, making it impossible to read from below due to its bad viewing angles.
The icons on the main buttons are ambiguous and need explanations on the left of the machine to be understandable.
In order to buy a single ticket, the user has to find the destination code in a long list of all destinations.
For daily tickets I let users select the zones manually, but offer to find the zone of a location, in case the user wants to go somewhere specific with their daily ticket. Additionally to the total number of zones, the summary screen also enables changing the day of validity of the daily ticket.
For single trip tickets I chose to let the user search for the destination they want to go to and display the zone in the results. This replaces looking up the destination in the list and typing in the number. The summary screen then shows how many zones the ticket is valid for to provide context for the price.
The first round of user test I conducted showed that the current design was not working well, despite all testers being not only able to buy the right tickets, but also the correct amount. While the users did buy the correct tickets, some of the information shown on the screens created more confusion than transparency to the flow, the biggest issue being confusion between what shows the amount of tickets and what the amount of zones.
The second iteration added more information to some of the screens where the users found them lacking. I decided to increase contrast to help them read the interface more quickly and redesigned some visuals to try and end the confusion between ticket amounts and zone numbers.
Testing this prototype showed that the solutions I came up with to solve the problems of the first prototype worked and the confusion was gone. The users were now able to buy single trip tickets and daily tickets quickly and reliably, even without knowing what zones their destination was in.
Design iterations over time
The last test showed that my prototype had eliminated the pain points of the current interface and thus speed up the purchase of tickets and eliminate the frustrations. For the transportation provider this can lead to lower rates of fare dodgers, be it purposeful or accidental ones, due to a quicker purchasing time and a clearer flow.
During research it became aparent that the current machines work badly for short customers and customers with bad eye sight due to the bad screen. As my interaction experience is mainly based on a screen, this screen has to be very good to not interfere with the experience. But the interface of these machines is not limited to the screen. The payment system of the machine and the physical appearance of the machine itself play a crucial role in ensure its usability. Thus making sure the touch screen is useable and implemented into the machine logically is very important to keep the user experience good.
I chose this project to get out of my comfort zone and work on an interface I don't use on a regular basis and its functional details I had to work out first. There is much more nuance in ticket machines than I originally expected, and designing a usable machine for all people is a much harder undertaking than it might appear. Furthermore, designing without any real conventions to lean on made each element in the interface a challenge to get right.
Tools used: Balsamiq - Sketch - InVision