Today I design FPGAs, engine control units and test systems for clients in automotive, industrial and medical technology — from my office in Alsace, right on the Franco-German border. The road here was neither straight nor comfortable.
Learning how to learn
The most important advice of my career did not come from a boss but from Professor Dr. Burhenne at Darmstadt University of Applied Sciences. He told us:
"The point is not to learn something — but to learn how to learn." — Prof. Dr. Burhenne, FH Darmstadt
At the time it sounded abstract. Today, after 35 years in an industry where technologies change fundamentally every few years, I know it was the most valuable sentence of my studies. It taught me not just to master the current state of the art but to look beyond — to recognise what matters today and what will matter tomorrow.
An early example: after completing my apprenticeship as a telecommunications technician at Deutsche Bundespost, they offered me a sponsored degree in telecommunications engineering. They refused to sponsor computer science. I chose computer science anyway — at my own expense. During my thesis I saw posters at the university where the very same Bundespost was desperately looking for computer scientists.
Sometimes you have to trust your instincts, even when the institution disagrees.
12 years in a corporation — what I learned
In 1990 I joined Mercedes-Benz in Stuttgart-Untertuerkheim. Engine development, rolling-road test benches, exhaust technology, later test software for engine ECUs. For a while I also worked in fuel-cell pre-development — a topic that fascinated me technically, although I got the impression that the wording of glossy brochures received more attention than actual technical progress.
Overall they were formative years — technically at the highest level, with colleagues who were just as passionate about the work as I was. Some nights we worked until three in the morning and security had to unlock the car park for us.
What I learned in the corporation that still helps me today
How to work within large organisations. How quality processes function — from requirements to series release. How to communicate with international teams. And how to write specifications that actually work.
But I also learned that technology is not always the priority in large organisations. I had superiors who were outstanding — and others who invested more energy in internal politics than in engineering. When the culture around me shifted in a direction where professional quality was no longer the deciding factor, I made a decision: I leave.
Jumping into deep water
In 2002 I founded my first company: Navimess Elektronik. What I had underestimated: the industry had an unwritten rule — "Nobody leaves Daimler voluntarily." Anyone who did was assumed to have something to hide. That did not make the early days any easier.
My first mistake was betting on public funding. Two years of applications, committee presentations, a positive expert opinion from a university — and in the end nothing materialised. The funds were spent elsewhere.
Lesson no. 1: contracts, not grants
Focus on customers, not on funding programmes. Every day you spend on bureaucracy is a day you are not winning a client.
How the first real contract came about
I did not get my first major contract through an application or a portal. I was on site at Audi for a testing assignment when someone asked whether I could develop CAN simulation hardware. Within a few days I had a working prototype — hand-built, but it ran.
That is the reality of freelancing in the embedded world: the best contracts arise when the client experiences your competence first-hand. Not on paper, not in a CV — but in the moment you solve a problem that is sitting on the table right now.
The certification question
For the Audi contract I needed ISO 9001 certification as a supplier. So I commissioned a certification body and went through the entire process.
Today, after more than 20 years of working with large certified companies, I see it more nuancedly. Certification tells you something about documented processes — but not necessarily about the quality of the work. I have seen certified companies deliver questionable results, and small engineering firms without certification that do excellent work.
Lesson no. 2: certification does not equal quality
If the client requires it, do it. But do not confuse it with a seal of quality. What counts is the work you deliver.
A fresh start in France
In 2012 I founded SCHMITT CONSULTING S.A.R.L. in Alsace. The proximity to the German border allows me to continue working for German clients with a lean company structure and minimal bureaucratic overhead.
My principle ever since: as few touchpoints with authorities as possible. No funding applications, no subsidies — full focus on what actually generates revenue: good work for good clients.
How to win contracts as an embedded freelancer
Honestly, most of my contracts came through recruitment agencies. That works well in the embedded industry because the large OEMs and tier-one suppliers often source through agencies. The downside: the margin that flows away can be considerable.
In the long run it pays to build your own channels as well: maintain direct contacts at clients, stand out through on-site competence, leverage recommendations from former clients. The CAN simulator at Audi did not come through an agency — it came because I was there and they saw what I could do.
Today I also use public procurement platforms and this website to become more independent of agencies.
What I know after 20 years of self-employment
Three things I tell every aspiring embedded freelancer
First: never stop learning — not just what is in demand right now, but what will be in demand in two years. From Pascal and assembler in the 1990s I progressed through C, LabView and CANoe to SystemVerilog, FPGA design and machine learning. Standing still is the death of a freelancer.
Second: your reputation is your capital. In the embedded world everyone knows everyone within two degrees. A satisfied client brings the next one. A dissatisfied one costs you three.
Third: only work for people who value your work. That sounds obvious, but it is not. As a freelancer you have the freedom to say no. Use it.