The ESA Call: On Being the Firm That Showed Up

In 2017 a small Polish company called Cervi Robotics won its first major contract with the European Space Agency. The job: build a drone that lands by itself, with no operator, and swaps its own battery. That company is now Dronehub, and that contract is where the whole thing actually started.

I want to tell this story straight, because it has been retold enough times that the edges have gotten shiny. The core is real and verifiable. Some of the lore around it is mine — my memory, my framing — and I will mark that clearly as I go. The lesson underneath is simple, and I believe it completely: the firm that wins is often just the one that showed up when the call was hard and everyone else stayed quiet.

What ESA actually asked us to build in 2017

The technical ask was specific, and that specificity is why it mattered. Not "make a cool drone." The brief was an autonomous drone that lands without a human at the controls and then swaps its own battery — so it can keep working without anyone walking out to the field to plug it in.

If you have never built flying hardware, that sentence sounds modest. It is not. Landing precisely, every time, on a fixed point — in wind, in changing light, on uneven ground — is a real controls problem. Then doing a mechanical battery swap reliably, outdoors, unattended, hundreds of times, is a second hard problem stacked on the first. Get either wrong and you do not have an inconvenience; you have a drone on its back in a field.

This is the genesis of the drone-in-a-box idea Dronehub is built on today: a docking station, a drone that services itself, and software that closes the loop so a person never has to be the limiting factor. In 2017 nobody handed us that architecture. The ESA brief forced the first version of it out of us.

The relationship is documented in ESA Space Solutions' business network, which is the part I can point to without caveat. It happened. It was about autonomous landing and self battery-swap. And it changed the company.

The part I tell as a story, and the part I can prove

Here is where I separate the two, because it matters for whether you should trust anything else I write.

The way I tell this origin — and I have told it on stages and in interviews — is that ESA reached out to a long list of European drone firms, something like fifty of them, and we were the only ones who answered the call. It is a good story. It is also founder lore. It is my recollection of how it felt and what I was told at the time, not a document I can put in front of you. So I keep it in the first person and I do not dress it up as a confirmed, independently audited fact. If a journalist asked me to prove the "fifty firms, only one answered" line with a primary source, I could not — and I would rather say that plainly than have it quoted back as gospel.

What I can stand behind: the 2017 contract was real, it was with the European Space Agency, and the value was roughly 200,000 euros. That is the number I have stated myself. I have seen bigger numbers attached to this story in places I did not write. I do not repeat them. The only figure I put my own name to is around 200,000.

I am this careful on purpose. When you are a young founder, the temptation is to let the impressive version harden into "the facts" because it helps you raise money and win deals. The problem is that one inflated number, once caught, poisons everything true you ever said. I would rather be vague-but-true than specific-but-false. The ESA relationship survives that test — it is one of Dronehub's genuinely verified European credentials, alongside our European Defence Agency and Horizon Europe work — which is exactly why I do not need to embellish it.

Why a 200K contract mattered more than the money

For a small team in 2017, roughly 200,000 euros is real money, but the money was not the point. The point was the signature.

A contract from the European Space Agency is a third party with no reason to flatter you saying: your autonomy thesis is worth funding, and we will hold you to delivering it. That signal does three things at once. It tells your engineers the hard direction is the right direction. It tells the next customer you are not two people and a slide deck. And it tells you, the founder, that the thing you have been insisting on against everyone's advice might actually be a company.

Before that, "autonomous drone that services itself" was a conviction I was defending. After it, it was a deliverable with a deadline and a counterparty. That shift — from belief to obligation — is what turns a project into a firm. You stop arguing about whether the hard thing is possible and start being measured on whether you shipped it.

It also set the pattern for how Dronehub would grow: through real European R&D, not vanity rounds. The same muscle that won the 2017 ESA contract later coordinated the Horizon 2020 project HUUVER and contributed to AUDROS — and I want to be precise here, because people conflate them. AUDROS is a separate, later feasibility study involving both ESA and the EDA, run by a multi-partner consortium. It is not the 2017 contract. Two different pieces of the record, both real, easy to blur if you are not careful.

From first prototype in 2017 to a real product by 2019

Winning the contract was the start of the work, not the end. We built our first working prototype in 2017 — the same year — and it was rough in the way all honest first prototypes are. It proved the loop could close. It did not prove it could close a thousand times in a row in the rain.

A fully working product was completed by around 2019. So the real timeline is roughly two years from first prototype to something I would actually stand behind in front of a customer. For autonomous hardware that has to land, dock, and service itself outdoors with no human in the loop, that is fast — but it is not the overnight version. There were two years of failed landings, mechanical jams on the battery swap, edge cases in weather and lighting we had not imagined, and the slow grind of making "it worked in the demo" into "it works every time."

I write about this honestly because the demo-to-product gap is where most hardware startups quietly die, and where I have made plenty of my own mistakes — I have written separately about what I got wrong over a decade in this field. The 2017 contract did not let us skip that gap. It funded us to walk through it. And the specific thing it forced us to solve — landing and battery swap without a human — is still the technical heart of the product. If you want the engineering tradeoff underneath it, I broke down battery swap versus charging in its own piece; the short version is that ESA pointed us at the swap path early, and that decision compounded for years.

What "answering the call" really requires

The reason this story is worth telling beyond Dronehub's own history is the lesson, and the lesson is not "be brave." It is more specific than that.

The opportunities that change a company are usually the ones phrased as a problem nobody wants. "Land with no operator and swap your own battery" was not a glamorous, obviously-winnable request in 2017. It was the hard, unsexy version of the work. That is precisely why the field around it was thin. Most people, most firms, route around the hardest version of a problem and toward the version they already know how to win. So the hard version sits there, uncontested, waiting for whoever is willing to say yes before they know how.

Saying yes before you know how is the actual skill. We did not have the autonomous-landing-and-swap system solved when we took the contract. We had a conviction that it was solvable and a willingness to be on the hook for it. The credibility came afterward, from doing the thing — not from describing it well, not from the size of the number, not from the story about fifty firms. From the drone, eventually, landing every time.

That is the part early founders get backwards. They wait to feel ready, or they chase the call that is already crowded because it feels safer. The call worth answering is usually the one that scares you a little and that everyone else is ignoring. Showing up to it — early, honestly, with your name attached to the deliverable — is a larger share of how this worked than any clever move I could claim.

Where I would start if I were you

If you are a founder or a student wondering how to find your own ESA call, here is the honest distillation.

First, go toward the hard, specific version of a problem, not the broad ambitious one. "Autonomous drone" is a wish. "Lands with no operator and swaps its own battery" is a contract. The narrower and harder the ask, the fewer people are competing to answer it.

Second, answer before you are ready, but be ruthlessly honest about what you have proven versus what you believe. Take the obligation. Then keep your claims tight while you go earn them — separate the lore from the facts in your own head, the way I keep "fifty firms" separate from "the 2017 contract was real and about 200,000 euros." That discipline is not modesty for its own sake. It is what lets people keep trusting you after the first impressive story.

Third, treat the credential as a starting line, not a trophy. The 2017 ESA contract did not make Dronehub. It funded the two years of unglamorous work — failed landings, jammed swaps, weather edge cases — that made Dronehub. The signature got us to the start of the real work. We still had to do the work.

That is the whole story, told as straight as I can tell it. The contract was real. It changed everything. And the reason we got it is the least sophisticated thing imaginable: when the call came for the hard thing, we picked up. If you want the longer arc of how I think about this work now, the about page has the full timeline — but the lesson fits in one line. Be the firm that shows up.