The danger of bad UX: the design that caused a Ski lift disaster in Georgia
My front-row experience of this chairlift disaster in Georgia and how it could have been avoided by practicing good design.
(I wrote this in April 2018, but didn’t get around to publishing it until now. Treat it as a live draft piece.)
Picture this: You’re on a snowboarding holiday in an exotic country with eight of your friends. It’s the last day of the week-long trip and you’re keen to make the most of it. Yesterday, the mountain was more or less closed due to weather. You wake up, pull back the curtains and see this:
You rendezvous with your friends in the kitchen, eat quickly, and get out onto the mountain as quickly as possible. The first run of the day is gorgeous, beautiful, sublime. The mountain is nearly empty. You charge downhill, hooting and hollering as you cut fresh turns in the foot of snow that fell the previous day.
As you head back up for another lap, you notice the one lift that’s been closed all week is open! It is the highest in the resort, tucked off to the side, but goes up to nearly 3300m above sea level. No one’s been up there all week. There are three huge open faces to ride, almost untracked, and all covered in fresh new snow.
You race over to the lift as soon as you can. There’s barely anyone there yet, so you get on nearly straight away. The second you get to the top, you strap in, and the whole group charges down.
It’s as good as you hoped! Views for days, no crowds and nice and steep.
You get to the bottom of the run and join the newly-formed queue at the base of the chairlift. After a while, you get to the front of the line and the chairlift stops. Out of frustration, people begin leaving. You’re stoked — you can be patient, then get the next ride down. No friends on a powder day!
Then, the lift stops. No big deal. You’ve seen this before.
You hear the auxiliary engine cut in, a slow whine that starts to build up. This mean that they’ll close the chair, let everyone off, and you’ll have to ride somewhere else today. These things happen — but it’s a bit annoying.
Then.
The lift starts rolling backwards.
⚠️ The video below may be upsetting to watch.
Slowly, at first, then faster and faster. You start to hear a strange noise, as the cable and chairs move backwards around the base. Then you hear people shouting and the sound like a train leaving a station as the chairs shoot backwards, out of control around the bull-wheel.
The gathered crowd in the line surges away from the out-of-control lift.
Screams.
Then a loud bang as a chair gets wedged sideways between the ground and the cable. There were people on it, and they’re tossed out onto the compacted snow and ice.
Bang.
Bang.
Bang.
More chairs slam into the wedged chair. A tangled mess builds up.
You don’t know it, but you’re shouting at people to jump off. Somewhere in the back of your mind, you know they don’t speak English. You keep shouting.
Somehow, after who knows how long — the lift grinds to a halt.
There’s a twisted mass of metal, and there are people in it.
That afternoon, you reflect that it was like witnessing a car crash — but if there had never been a car crash, ever, anywhere.
Days later, the investigation returned by Bureau Veritas (by request of the Georgian government) had some startling results:
“…after the chairlift was stopped, the operator had to introduce specific sequence of procedures …Unfortunately, according to the current conclusion, the operator made a mistake. The combination of the actions that he should have had carried out were not implemented in compliance with the relevant instructions — it was a human error.”
How did a situation like this happen?
This is where, as a designer, I have to step in and say no. I design products and systems for people at work — and this just isn’t good enough.
Years ago, I worked on chairlifts at Thredbo Alpine Resort in New South Wales, Australia. I’ve worked in an industrial facility, too — and in a situation this critical, operator error doesn’t cut it. The design of the system and infrastructure around it is the problem here.
So let’s explore the factors that led to this happening, and some recommendations on how to prevent it:
What does the interface for these chairlifts look like — and how could it be better?
The first (and most obvious) thing is to blame this on the interface. A chairlift operating interface looks pretty similar to other industrial control surfaces: big clear physical buttons, small jargon-y labels and maybe a couple of indicators. Inside the base terminal, there’s sometimes a skeuomorphic diagram layout of the physical shape of the stations with indicator lights.
If you’re an interface designer, these might make you wince. They can definitely be improved. Looking at these, you can see how an operator might complete the wrong ‘sequence of actions’.
Suggestions:
- Be consistent in the layout
- Use different control types for different function. even something as simple as different rotary dial shapes
- Align labels more closely to each control — and use bigger, clearer labels.
- Group common functionality in logical regions.
How does the user flow allow a disaster like this to happen?
Regardless of how consistent or practical a layout design is, it’s critical to map out every exception in the user flow. As a designer, you have to know what will happen when someone pushes all the buttons in any order. Even if it’s an edge case, a designer should understand the implication of a user not ‘doing the right thing’
It’s pretty clear to me that the exception flow for chairlift operations still allows a user to make this critical mistake. For this type of chairlift (Doppelmayr fixed-grip quad), the operator has to follow this sequence — and if they make a mistake there’s a catastrophic consequence.
Suggestions
- Map out all exceptions — and understand how to resolve them. What happens if a user gets ‘stuck’? In this kind of environment, they need to be able to go back.
How did the service allow this to happen?
Leaving interface and flow behind, the biggest issue isn’t that an employee made a mistake — the biggest issue is that it was even possible for this mistake to be made. While there are glaring flaws in the design of chairlift operating panels, the training and implementation of this service is culpable too. It’s clear the employee in question here was not sufficiently trained to operate the (albeit awkward) system he was using.
Suggestions
- Technology doesn’t exist in a void — understand that there are real humans under real pressure using the systems.
- Test, test, and test again.
Why wasn’t there investment to make the employee tools better?
As designers, it’s our responsibility to keep the wellbeing and safety of our users paramount. In digital product design and service design it’s not often dangerous to life and limb but:
If your front-line employees can’t use their tools effectively (digital or otherwise) then your organisation is setting itself up for failure.
For a comparatively small ski resort in a developing nation like Georgia, it’s understandable that there are larger geopolitical and economic issues at play in the implementation of good design practice. A significant culture change could be in order, as mountain resorts around the world will use this incident as a compelling reason to reform their operations.
There are probably thousands of poorly designed interfaces in the world. Not just in ski resorts, but on factory floors, assembly lines, heavy equipment and more. If you find yourself working on this kind of project — be aware of the people and the environment you’re designing for. Sometimes, bad design changes peoples lives — forever.
Thanks for reading. Mash the 👏 button to share this article with others.
Ed Carroll is a designer currently managing digital service at Vodafone in London. You can follow him for more articles, or visit edcarroll.design to learn more.
