MAY 29, 2023

BRAID’S TEAM

<aside> 💡 Welcome to the fourth article in our series of Tech Team Profiles! Here, we’ll introduce you to the team building Braid’s automated engineering design technology. Sjors Donkers is a Software Engineer at Braid with a background in geomatics and autonomous driving.

</aside>

Briefly describe your career in tech thus far.

I completed a bachelor's degree in Aerospace Engineering and a minor in Computer Science at Delft University of Technology in the Netherlands. After, I completed my master’s degree in Geomatics, which involved mapping, computer vision, and geometry processing. My specialization was standardization of data formats and validation for geometry processing.

I worked as a consultant and engineer with Sweco in the Netherlands. The software that I helped develop was somewhat like “Photoshop for maps.” We acquired and maintained geodata that allowed us to have centimeter-level accuracy of everything you can see in the Netherlands that was up-to-date within three months of changes. High resolution data are important for cities implementing urban development projects, as well as individuals who may be starting construction projects and need detailed information about their property.

Sjors marveling in the beauty of Mt. Fuji in autumn.

The Netherlands as a country has a lot of publicly-available data and many companies collaborating towards this effort. Although I was primarily a software developer with Sweco, I would meet with our customers typically twice per week to support and train them with our software. As a result, we would incorporate their feedback to improve the software. In that sense, it was a rapid and user-focused development effort.

Eventually I came to Japan and joined Woven by Toyota, which was previously called Toyota Research Institute for Advanced Development (TRI-AD). I was the first engineer to join Woven from outside of Japan. We strived to apply the Silicon Valley-style development process to Japanese craftsmanship. I started on the mapping team for their autonomous driving system that would eventually be known as Teammate Advanced Drive. Our team was responsible for taking in maps and ensuring that they would work within all of the other embedded applications. We also did all of the data pipelining, validation, and checks to confirm that the car is safe to drive. It was a safety-critical development environment.

We started as a small team that grew to about 60 people. As the Tech Lead, I would provide general technical direction, launch new projects, as well as assist with hiring. I’ll admit that I can be a bit idealistic: if I see opportunities for improvement, I will provide that feedback to my team and management. This philosophy about work ultimately led me to become the employee representative, which meant working with HR to voice employees’ concerns. Ultimately, I am very proud of what my team delivered, especially given the circumstances.

What initially sparked your interest in tech and geomatics?

Basically, I knew that I wanted to be a nerd. High school students in the Netherlands are required to choose a specialization. Since I performed the best in technical classes, I chose the technical track. In university, I chose to study aerospace engineering because of its practicality: it produces results that you can see. I found that I was good at programming and then pursued geomatics. Like aerospace, software engineering is a field that produces tangible results, though they are realized on a much faster timeline. I enjoy the near-immediate feedback of software.

The first introduction I had to geomatics was honestly the most boring presentation I’ve ever seen, but there were aspects about the field that I found- and still find- appealing. I am interested in the fact that working with geodata must be performant given the size of the dataset. Additionally, unlike other kinds of data, it’s much more challenging to identify the edge cases in the dataset. In short, my brain just works well with this kind of data. The natural affinity between my background in mapping and Japan’s automotive industry is what eventually brought me to Japan.

Sjors enjoying a bit of French culture outside Notre Dame Cathedral.

Do you have any open-source projects, geomatics-focused or otherwise?

The focus of my master’s research was connecting the architectural world with that of GIS. The tool allows you to extract architectural data into city level data. The code from my master’s thesis is surprisingly still in use; just recently I was recognized at the Tokyo Rust Meetup as a result of that project. Separately, I have several private geometry processing libraries, one of which I will present at the Tokyo Rust Meetup in July.

Any other interesting hobbies and activities in Tokyo?

I’m a big Formula 1 fan. I’m Dutch, so I’m obligated to say that Max Verstappen is my favorite, but I am secretly cheering for the other Dutch racer, Nyck de Vries. What makes F1 special compared to other race categories is the importance of rapid technical innovation; because the car design is not fully prescribed by race regulations, there are many strategies and tricks to improve the car’s speed every race.

Also, I am very interested in emerging programming languages and their use cases. For instance, I’ve gone down the rabbit hole that is the Jai programming language from Jonathan Blow. Other than that, my favorite activities around Tokyo are barbecuing at the beach in Odaiba and going to karaoke with my wife and friends.

Sjors and his wife spending time at a family farm in the Philippines.

After working heavily with geodata and in autonomous driving, what led you to make the leap to working with Braid?