Sebastian Thiemt is a Mechanical Engineering graduate specializing in Industrial Design. In 2015, as part of an innovation competition at the internal "Digital Life Day", he and two colleagues came up with an idea for an app. There is now an entire team of eleven people responsible for developing the "pacTris" app. pacTris uses artificial intelligence (AI) to help customers load their vehicles efficiently. In his interview, Sebastian explains the app's unique features, and tells us how an idea grew into a team.
Hello Sebastian. You joined Daimler in 2012 as a trainee in an international talent program. But that wasn't the first time you came into contact with Daimler, was it?
That's right: as a student I completed an internship with Daimler in Palo Alto in the US. To be honest, before then I thought that almost everything at Daimler revolved around technology and mechanics, and not necessarily digitization, AI or apps. But that industrial internship at RDNA (Research & Development North America) showed me the error of my ways. We conducted research and development as part of an international team. Among other things, we programmed one of the first Mercedes apps. That is where my affinity for all things digital first flourished, and that time ultimately laid the groundwork for our "pacTris" app.
What was so surprising at Daimler?
Firstly it was the capacity for innovation, and the search and desire for the best solutions. But it was also the work culture in which people from a wide range of expertise can form a great team, and how a heterogeneous group can be motivated. That is something I experienced there, and I learned a lot for my current work group.
Let's talk about your "pacTris" app. How did it all start?
Two colleagues, who I got to know on the talent program, and I wanted to take part in the first Digital Life Day in 2015. The innovation competition's motto was "Making technology tangible for customers". We had the idea of developing an app with artificial intelligence to tell customers at the DIY or furniture store whether what they want to buy fits into their car, and the best way for them to load it. This is actually a typical "knapsack problem", i.e. what is the best way to pack different objects into a backpack, case or trunk.
Before then I thought that almost everything at Daimler revolved around technology and mechanics, and not necessarily digitization, AI or apps.
What gave you the idea?
It's probably the fact that I have moved house so many times – with a very small car. For my first house move I wrote an Excel program that calculated the most efficient way for me to loan my car. But that program was of course completely inflexible, and it is of course also highly impractical to stand in a furniture store with an Excel table.
You won with your idea, is that correct?
Yes, that was fantastic, and marked the start of our development phase. It all started with "pacTris" for smart. But our dream at the time was to support all Mercedes-Benz models. And it ultimately worked. There are now eleven of us on the team, and we will successively offer pactris for all Mercedes-Benz model series. The AI algorithm has changed quite a bit over the years. We now have AI developers on the team, and there's not much left of the original algorithm [laughs].
A report in the intranet then brought you to the attention of a logistics planner at Daimler…
That's right. And now we are conducting pilot trials of "pacTris pro". This is an app specifically for logistics applications, and not just for planners, but rather for the forklift drivers who have to load the containers. I now also have a forklift truck license because I wanted to understand the world of operating logistics and not simply design a product for something I never did in my life. The app responds to deviations within seconds. This means that if a consignment has not arrived or was delivered in a larger package than planned, the app registers this immediately and carries out a new calculation for the forklift driver within a few seconds. The special thing about all of this is that the system is completely flexible. A comparable product on the market would take hours to carry out a new calculation.
For my first house move I wrote an Excel program that calculated the most efficient way for me to loan my car.
How does your app do it?
The success of "pacTris pro" is actually based on three features: Firstly, we have a very clever data crawler, secondly, our AI combines two different technologies (namely genetic algorithms and "reinforcement learning"), and thirdly, we combine all of this with innovative 3D sensors.
We are currently trialling the system in a large logistics hall near Speyer. The results are very promising, and have surpassed our expectations. Our app has reduced the use of entire shiploads of containers! We are already receiving inquiries from other parts of the group.
That sounds really great, and we are interested to see where we will encounter "pactris" in the future. Can you tell us what you like about working for Daimler?
The opportunities! The fact that the group combines so many different specialist units, and the access to a large network of experts, is certainly unique!
I now also have a forklift truck license because I wanted to understand the world of operating logistics and not simply design a product for something I never did in my life.
And one personal question to finish: It's 10:30 a.m. on a Saturday. Where are you usually to be found?
Honestly? You'll probably find me at the DIY store. I'm an engineer at heart, and I usually start a small construction project every weekend! I always go with my young son – he loves DIY stores too!
Sebastian Thiemt is 32 years old and a Mechanical Engineer specializing in Industrial Design. The father of a two-year-old son studied in Stuttgart, Germany. Sebastian's first experience of work was developing chainsaws for a tool company. He will not quickly forget his first training module as part of his training program at Daimler. It was 2014, and together they watched the semi-final of the soccer world cup between Brazil and Germany. Germany, who went on to become world champions, won 7:1. Interestingly, his second training module three months later was held in Brazil.