While he may not be in love with parking, he does love solving problems, which is why Oliver Buterbaugh developed the app, carparks.je. He tells Natasha Baluchi how it happened
FOR many drivers, finding a parking space in St Helier can be a source of frustration, particularly during the school holidays or in the run-up to Christmas.
For 22-year-old Oliver Buterbaugh, it became a problem worth solving.
What started as a university project for his degree has since developed into a fully formed app, showing live availability across seven state-owned car parks in St Helier, while forecasting how full each one is likely to be when drivers arrive.
Having already been downloaded more than 2,000 times since its launch in July and reaching number 34 in the navigation category on the App Store within hours of going live, Oliver hopes to continue to expand the service to further help his community.
“I always wanted to do create some sort of AI model from scratch, that would help Jersey,” Oliver said.
“I ended up doing that in parking, which I know doesn’t sound like the most exciting thing but it’s useful, and often the useful things are quite boring.”
Born and partly raised in Jersey, Oliver spent some of his childhood years in Florida, before returning to the Island and studying at Hautlieu school for his A-levels.
After finishing school, he went to London where he studied computer science at university, and it was during that course that the original idea for his app materialised.
“It was for my final year thesis that I started this with a very scaled-back version,” he explained.
The choice of Jersey was deliberate, as he said he wanted to choose something that was “close to home”.
His starting point was the government website.
“I looked on the Government Open Data Programme, which has great data sets and I’m very impressed with the Island and government for doing that; they deserve a lot of credit for that,” he said.
“And then I found the availability stream of parking data.”
Having found the availability stream, Oliver realised he could do much more than simply show how many parking spaces were available.
From there, he was adamant that he could begin to identify patterns and use them to predict what parking might look like in the future.
The idea felt easy for Oliver at the time: a driver could open the app before their journey, select where they would like to go and see how many spaces would be available.
Building it, however, proved more complicated.
“Honestly, going from the idea to the prototype only took a couple of hours,” Oliver said. “But because the government doesn’t have any historical data, I had to just patiently wait and have it recording the data in the background.”
The original university mock-up collected around two and a half months worth of information.
The turning point came when Oliver’s sister, who had the original version of the app on her phone, started using it frequently and asked when he was going to update it.
“When my sister was actually using that first version it and wanted me to update it, that’s when I thought about taking it seriously,” Oliver recalled.
He then returned to the project more than a year after submitting his thesis, and decided to rewrite it from scratch, hoping to make a “better” version.
While the initial idea was developed quite quickly, he estimated that the current version took around six months to develop, with one month being spent polishing the product to make it more user-friendly.
The difference between the university project and the current app is its ability to respond in real time.
The original idea relied on a “frozen snapshot” of parking patterns, which meant it could not account for unexpected events affecting demand in town.
“All it knew was that, for example, on a Tuesday, there are usually this many spaces here.
However, now, if an event unexpectedly affects parking demands, the system can look back at previous occasions,” explained Oliver. “It takes the parking that it saw during other events, shapes it to match what it’s seeing right now and is able to properly react in real time.”
In addition to this, Oliver designed the app to prevent drivers from having to check the government website.
“The whole point of the app is you open it, you tap where, for example, you work, and it knows how long it’s going to take you to drive to town, it puts in a little buffer of five minutes and it says, ‘this is how many spaces are going to be there’,” he said.
“This stops people getting to the first traffic light on the avenue, pulling up the website when they are driving, which they shouldn’t be doing, and seeing where to park.”
Oliver says the app is currently 98% accurate when predicting availability one hour ahead, based on a rolling seven-day period. At the time of the interview, the maximum difference between the predicted and actual number of spaces was 13.
He considers one hour ahead a useful benchmark, saying it gives drivers enough time to travel into town, while the predictions become increasingly accurate as the arrival time gets closer.
The system, he adds, is also continuing to learn. Rather than having to account for seasonal changes in parking patterns, the system has the ability to identify the changes itself as more information is constantly updated.
As the app only has access to seven state-owned car parks in St Helier, Oliver hopes to extend his ambitions beyond its existing coverage and connect other car parks in Jersey, which are not already part of the government’s availability stream.
And Jersey is only the beginning.
Oliver is already working on a “prediction pipeline” with 6,000 car parks around the globe ingesting into his system.“Luxembourg is my next ambition as it is quite similar to Jersey,” he said. “I want to see if I can penetrate a different market.”
Moreover, he said one of the most encouraging parts of user response had been seeing people defend the app online.
“There was one comment online that was clearly intended to be bad, but there were 60 or 70 responses that were saying to just try it out,” he said.
“When you are defended online, especially on Jersey Facebook pages, you have made something that people actually agree with and like.”
The community response has also reinforced his belief that Jersey is a great starting point for local developers to experiment with new technology, with Oliver believing that the Island’s open data could provide opportunities for other people to build tools around real-world problems.
“If you have an idea, you can open up AI, like Claud for example, and say, ‘build me a tool that does this’ and find the data stream.”
The young developer sees AI as a means “to help”, and ultimately make software development more accessible, rather than simply replacing existing jobs.
Despite his international ambitions, some of Oliver’s proudest moments have been much closer to home.
“My grandma texted me saying that her hairdresser’s parents were using the app,” he said.
“Little things like that make me very happy.”
With ideas about new features already on the table, including bringing the app to Android and eventually Apple CarPlay, Oliver understands that there is a long way to go.
“I genuinely do care. I hope it comes across in the app,” he said.
While aiming for the app to become a daily habit of lots of drivers globally, Oliver said his focus remained on creating technology that was useful rather than simply being impressive.
“I do care about making things that are functional, useful and maybe a bit beautiful,” he said.
And parking is only the beginning.
“I’m not in love with parking, but I’m in love with solving problems, and this happened to be the one.”
