Exit to DJI, working at Tesla and building in Y Combinator. Meet Timothée.
During EPFL, Timothée built MotionPilot, an FPV drone controller which he sold to DJI. He then joined Tesla for 2 years. Currently, he's at YC building atopile, a language to describe circuit boards.
Studies MSc Micro-Engineering @ EPFL ‘18, BSc Micro-Engineering @EPFL ‘16
Experiences Tesla Battery Design, Tesla 25 Guns, DJI, Green Motion S.A.
Companies atopile (YC W24), MotionPilot (acq. by DJI)
Origin Geneva
Links LinkedIn
Growing up building things in Geneva
Where did you grow up and what projects did you work on as a kid?
I grew up in the countryside of Geneva. As a kid, my dad brought me some basic electronics early on, setup on a piece of wood with transistors, LEDs and light bulbs that I could connect a battery to. That’s how I first got into engineering.
One thing we did a lot with my dad was to build remote controlled aircraft. The first one was a large glider that we built in the basement. This taught me how to plan a project and use my hands to build it out. Building flying objects was always very exciting.
Later when I was around 14 or 15, FPV (*First-person view for drones) started becoming a thing. People would take wireless baby monitors and put the transmitter on an RC plane and the receiver on a small TV. That way they could fly from the drone’s perspective. There was one guy who did this and started a company in Morges called ImmersionRC, so I purchased some of his equipment, threw it on one of my planes and started flying FPV around the neighborhood. It started growing pretty quickly from there and I began building drones as well.
Moving to EPFL
Tell me about your furniture project, Sark.
Sark started when I moved to Lausanne to attend EPFL. My thought was that instead of just going to IKEA like most people do, I wanted to build my own furniture. The first project was a shelf I built using threaded rods and pieces of wood. The shelf turned out pretty good so then I started making lamps as well.
The experience culminated when the three of us roommates, who all loved music, decided to build a sound system. I came up with a pyramid-shaped subwoofer and put the tweeters and woofers into two towers. It was a whole adventure building your own furniture and sound systems.
As I did these projects, people showed a lot of excitement about them and wanted to purchase them. So I ended up doing small production runs - now there are maybe 10 people in Europe who have a version of my lamp or shelf.
What made you choose EPFL and Micro-engineering?
Everyone is a doctor in my direct family - both my parents and my two older sisters. So for me, the logical path would have been to follow that route too. But I always wanted to do something different. Since I was a kid, I was always building stuff, so EPFL seemed the way to go.
One of my older sisters actually started at EPFL in life sciences before becoming a doctor. So through her, I heard about what it was like to be at EPFL and it seemed exciting. That helped me decide to pursue it.
Did you ever face self-doubt or imposter syndrome at EPFL?
Oh yeah, the first two years at EPFL were pretty intense. I didn’t really have time to do other projects, apart from maybe the furniture.
As time went on, I started learning about topics I wasn’t familiar with, like electronics and microcontrollers. That was exciting since I had been using those device for years and finally understood how they worked. Being surrounded by people in the field also helped a lot to dive deeper and learn about the subject. I remember having a PhD student teach me how to use Altium to design circuit boards for example.
I started learning a lot and it became obvious that I wanted to build something in that space. My friend Thibaut was excited about that too and we started building a remote control for drones that worked with hand motion. Pretty quickly, we ran into firmware and electronics issues, that’s when our friend Arthur and Benjamin joined in and helped develop the project further.
As we kept building, people started getting excited. Ultimately, it snowballed into a small startup that we were running alongside our studies.
Building MotionPilot, a Motion Controller for FPV Drones
How did MotionPilot takeoff?
MotionPilot started after classes and during weekends. The first prototypes were built with Arduinos and Adafruit modules. Once we had a working prototype, we showed it to one of our professors and he offered us to continue building it in his lab. After winning a few startup competitions, we were able to rent our own offices next to EPFL.
The initial idea was to build a motion remote only. But after discussing with professors at EPFL, we started looking into adding haptic feedback so the user could feel in his hand the motion of the drone. This ended up being a rabbit hole. We spent years developing all kinds of prototypes that would create vibrations or various types of pressure in your hand.
Did you ever consider quitting your studies to focus on MotionPilot full-time?
At that point, it was still very much a side project. We didn’t feel at all that it was time to stop everything and just work on that. I’m glad we continued focusing on the studies since there were some amazing classes that I’m super happy I went through. For example a class on semiconductors and their application in industry.
I also didn’t feel like MotionPilot was taking time away from my studies. Maybe it was the specific combination of classes I took, but I still had time on the side so I wasn’t really in a situation where it was a tradeoff.
One thing we did was to negotiate and have our master’s project and internship be allocated to working on our own startup. We had to ask a few times but we ultimately received approval to move ahead. I’m very grateful for that to EPFL.
I believe they’ve turned this process into a recurring program where students can spend a semester focusing on building their own technology and venture.
The MotionPilot journey
How did you fund the early days of MotionPilot?
We won a few startup competitions - the Xgrant which granted us with 10k CHF, Venture Kick with 130k CHF and the START Contest, organized by START Lausanne, which granted us 30K CHF.
What were some of the lowest points in the MotionPilot journey?
There were two low points:
The first one was when we decided to abandon haptic feedback to focus solely on motion control for the remote. At that point, haptic feedback was a core aspect of the design and abandoning it felt like abandoning the whole project. In retrospect, this was a great decision. It enabled us to focus on other aspects of the controller, like the motion control algorithm and the long range communication protocol. It’s through this process of focus that the controller became a great product.
The second low point was towards the end of the project. We had spent years building the haptic feedback mechanism and had recently pivoted to doing only motion control. We were running low on funds and didn’t have a clear path forward to market. The situation was unlocked when two people entered the story, Estelle and Ramy. Estelle and Ramy saw value in something we hadn’t considered before: filmmaking. With them, we met the best drone filmmakers across Europe, which ultimately led us to meeting and working with DJI.
Working at DJI
How did the DJI technology acquisition come about? Did you reach out to them?
The one thing we did wrong with our project is that we didn’t launch enough iterations of our product which meant it took too much time to figure out what people wanted. For example, we spent years developing iterations of our haptic feedback mechanism, without testing it on our users and confirming that was the most valuable aspect they were excited about.
When Estelle and Ramy came into the picture, they saw the MotionPilot controller as a professional tool for filmmaking. From that perspective, they were extremely excited. So we made a plan to talk to everyone in the FPV drone filmmaking industry that we could get in touch with.
We met many of the best FPV pilots in Europe - we drove to France, Germany, Austria, and across Switzerland. Through them, we got introduced to different companies. Ultimately, one pilot introduced us to DJI’s CEO. So we flew to Shenzhen with our product and met the DJI team in a park. DJI’s CEO started flying with the remote and crashed on the first flight. I was quite stressed at that moment. It turns out that it was due to the goggles failing in flight. So he took off again and he loved the flight experience. He very quickly decided we should proceed with a collaboration. The timing was perfect because they were developing an FPV drone at the time and they were not satisfied with the remote that was paired with it. MotionPilot seemed like a great alternative.
How did you go about negotiating with DJI?
Well, everything started in a park in Shenzhen like I mentioned before. This is where the decision was taken to collaborate. 20 minutes later we were sitting at the DJI headquarters and discussed the details of the collaboration.
This whole process was started extremely fast compared to other companies we had interacted with, which was refreshing to see. After that it did take a few more months to iron out all the smaller details in the agreement.
An advice I would keep to myself is that negotiating with those large companies takes a lot of time. There is an enormous amount of inertia in entities of that size and trying to change some minute details can take months to figure out. I wasn’t expecting how consuming this whole process would end up being.
There were some other smaller companies that were interested in working with us. With them, the process would have been way faster but the impact probably not as large as what it turned out to be with DJI.
Working in Shenzhen
What was it like working in Shenzhen for DJI after the acquisition?
It was very interesting. First of all, I had lived most of my life in Europe, so entering a new country, a new culture, and having to adapt was a big change.
When I initially arrived at DJI, I wasn’t immediately allowed in the building. The first meetings happened at the local Starbucks. An engineer would come down with a few ideas, we would chat about it and they would go back to the main building. It took me a few weeks to gain the trust of the engineering team until I was allowed into the building.
I put a lot of effort in helping the team and ultimately got put in charge of executing the design. We went through a few iterations of the controller shape, placement of the buttons, flight modes and control algorithms. This whole process took about 6 months.
Joining 25 guns, Tesla and meeting Elon Musk
What drew you to working at Tesla in the 25 Guns team after wrapping up with DJI?
First of all, the engineering being done at Tesla was super exciting and the speed at which they were executing on really complex products was impressive.
At that time, it wasn’t exactly clear to me where we needed to go as a society to improve the climate situation. So I was curious to discover whether electrification would be one solution to the problem. Working in that field seemed like the logical thing to do to figure it out.
Did you get to meet Elon Musk while working at Tesla?
Yeah, when I was a 25 gun engineer, we would have meetings with him. I never met him in person because when I was in the US, he was in Europe and vice versa. But the rest of the team did!
How did you decide to leave Tesla to start a new company? What ideas were you considering?
Already when I joined Tesla I had ideas in the back of my mind. There were three main topics I was excited about:
Foiling boats to transport merchandise. Something that could serve a need between really expensive planes and very slow cargo ships.
Agrivoltaics - putting solar panels above crops to simultaneously grow plants and generate electricity.
Electric bicycles as a replacement for cars in our transportation infrastructure.
I researched those three fields and ended up starting a new project related to electric bicycles. The goal was to make an open source gear-less electric bicycle drivetrain. I worked on this with a few friends during a couple of years.
Founding atopile and getting into YC
How did you begin your new company atopile, and why did you decide to apply to Y Combinator?
As soon as I stopped working on MotionPilot, I started working with a few friends on a new drivetrain for electric bicycles.
As I was working on that, I realized that the electric bicycle market was an extremely crowded space - everyone was building e-bikes. So the question became, what could be different about this project? How could we build it in a unique way?
That is when open source entered the picture. What if we could build a piece of hardware but collaborate and share it like we can share software? We started building early versions of the project and we would version control the KiCAD files on GitLab. I continued working on the project with some colleagues at Tesla when I moved to the US.
Through this process, I met Matt and we started talking about how to make open source hardware work and how to collaborate around hardware design.
He had been thinking of ideas to build “git for hardware” similar to what git enables for software in terms of collaboration and version control.
We discussed the problems that existing companies face and how the fact that hardware design tools are so far behind software tools is one of the reasons it’s so difficult to develop hardware products, especially in large companies.
At first it was just discussions, then we started spending our evenings developing quick prototypes of what a solution might look like. We went through three iterations of the idea until we ended up designing our own language, which is the core of what we’re working on today. As we were building the tool, Narayan, another friend from Tesla started using it more and more and ultimately decided to join the project too.
We were working on the idea on the side for a little while, and then came the consequential decision point of leaving our jobs to focus on it full-time. We considered if we could bootstrap the company but our visa situation didn’t allow it.
That’s when we considered joining YC. We applied and got accepted. From there, it was a pretty straightforward decision that it was an opportunity we could not miss.
A few days ago, we launched atopile on hacker news and have seen a lot of excitement since then!
General questions and advice for students
What advice would you give yourself if you were back studying at EPFL again?
I would spend as much time as I can with engineering student teams. There are a few that work on various race cars, rocket, boats and other cool projects. When I was at EPFL, I did one project with the Hydrocontest racing team which was great. Those are super exciting people to be around with.
They recently setup a Makerspace at EPFL which I think is awesome. I would probably spend all my time there and meet the people who work there.
Within those groups, there are always a couple of people who are excited to turn their project into a business too. I had some friends I met through the Hydrocontest who are now building a racing boat and they ultimately want to spin the project into an engineering consulting company.
I think it’s also good to work on something that isn’t necessarily a business with people first, so you can get to know them and see how they work. Then if you’re excited to keep working together, you can embark on a more ambitious project.
Should aspiring entrepreneurs in Switzerland try to get to the US as fast as possible?
This one is tough for me to answer. In an ideal world, you wouldn’t have to leave your home to go where your field is most active. Unfortunately that’s kind of how things are currently even though it seems the trend is inverting.
I would say yes and no. If you’re in a very specific field and you know you want to work with the best people in that field, there is usually one place in the world where the experts in that area congregate.
If you want to master watchmaking, being in Jura is probably the right place because that’s where the knowledge is concentrated. If you want to be on the cutting edge of software and IC design, then the Bay Area is the hub.
If you want to be with the best people in your specific field, go to that place. On the other hand, there are lots of opportunities to build successful companies wherever you are.
When I was studying at EPFL, I used to work one day per week at Greenmotion. They were becoming the largest installer of electric car charger in Switzerland and they have done great. Overall, I’m very impressed by people who stay in their home country and build great technology that benefits the people living around them. We need more of that.
What was the most valuable part of your EPFL experience?
The people I met. Arthur, Ben, Thibaut, Estelle and Ramy who I ended up spending years working with on MotionPilot - I would not have had the opportunity to meet them if I hadn’t gone to EPFL.
Same thing for the professors and coaches who helped us in developing our startup by providing time, money, and resources to make it happen.
Closing notes
Hope you enjoyed this week! Speaking to Timothée was incredibly inspiring - someone I never would have met if I hadn’t started the newsletter.
There are some super inspiring people out there at EPFL, ETHZ, TUM, everywhere. Probably the best place to be surrounded by them is in the engineering associations and entrepreneurship/co-working spaces!
As always, message me on LinkedIn your thoughts on who I should feature next, what you want to ask the guests (more studies-focused or industry-focused?), what deeper questions you’d like to ask!