You don’t have to be a data scientist to indulge your curiosity using analytics. Case in point: German medical student David Häsler, who started programming his way to better parking options in the city of Heidelberg as a personal side project, which eventually landed him on a SAS Hackathon team focused on better mobility through the city overall.

The team’s goal was to analyze intersecting mobility data – traffic, bikes, people, public transport and parking – with weather and event data to create forecasts aimed at improving urban quality of life and sustainability.

The team’s mentor described Häsler as becoming a “black belt” of the SAS® Viya® platform in the four weeks of the Hackathon despite having never used it before. So obviously, we were eager to hear how it all came about for the 21-year-old future physician.

How much did you struggle to find a parking spot that you began looking into this in the first place?

I don't even own a car! Virtually no student here does. But I live in the city center and parking is very scarce there, and every time my parents visit, we have to look for a spot and it's expensive. From the beginning, I thought if this application is able to scrape the API the public website offers and I could accumulate the data over a certain time, I could train models on predicting the available spaces. So I thought it would be great if you could take out a bit of a hassle there for everyone.

So you already had a bit of a technical background?

I did computer science in high school, but I was missing being able to apply it to a real project. I spotted the opportunity with this parking website right in my hometown. Also in medicine, I’m interested in radiology, and it’s becoming more and more data driven. The models are increasingly able to achieve similar results to humans, and I found that quite stunning and a great future for the medical field.

So what parts of the project specifically did you work on?

The main aspect from the Hackathon was the prediction feature. The results were two different approaches – a long-term prediction and a short-term prediction. For every parking garage you can select the date, and then it will predict the occupancy for this date in the future with the time and the available spots. If you're looking at the occupancy in the short term, there are other factors to be determined, because maybe there's a special event that didn’t happen on that day last year. It's not perfect and can still be improved on, but I managed to bring this into production, and the other one is in a good spot.

Even though you had coding experience, how much easier was it having low- and no-code options?

To really understand what the model builder selected for you and maybe to optimally work with it, I think it's also essential to have a deeper understanding and maybe be able to do it in the classical programming interface as well. But I think it's possible to develop working models which are sufficient for most use cases with a lot less entry barriers then there may have been before.

What were the biggest advantages of being able to use SAS Viya to work on your project?

You can achieve things very quickly that wouldn't be possible with the open source tools I used to develop the Python (parking) application. You could easily prepare animations which showed the occupancy throughout the day, and that gave me a good starting point to iterate further and develop the models. Obviously, it was exciting that you could auto train the models … so it was possible to develop comparable models which outperformed the ones I developed with the open source tools in a comparatively short time.

Your team’s Hackathon mentor described you as a “black belt” of the platform despite no prior experience with it. What was it like jumping into that on this project?

I already could program, and having a technical understanding in general was definitely helpful, but also it was an awesome experience being on the team. Everybody worked together and could ask questions, and I think it elevated my knowledge like a lot quicker than it would have been possible otherwise. There was a lot of experience on the team, and I was very keen on helping.