Elias

Four years ago, I somehow decided that I want to learn how aircraft take off from water. It seemed fascinating to me how first the forces of water, and then the forces of air enable it to fly off the water. I had a pretty good understanding of the air part from flying gliders and radio-controlled model airplanes and from working together with an Aerospace engineer for two years. But I had no clue about the water part. All I knew was that water is more than 800 times more dense than water.

So I found a company that had designed a bold new electric seaplane and seemed to be struggling a bit. They had built a prototype and flown it from a runway, but there was no video of it taking off from water.

I found the number of the founder on their website and wanted to call him, but kind of forgot about it. Eventually I felt quite sad and depressed for no apparent reason. it was almost the middle of the summer. In a moment of inspiration I asked myself:
“What is the one thing that I'm afraid of that, if I did it, would make the whole day a success?”
It was immediately clear to me that I had to make this phone call. Then I imagined the worst and best possible outcomes of making the call, and it suddenly it was very easy.
There is an important lesson hiding in plain sight here: Whenever an action has a clearly favorable asymmetry of worst and best outcomes, it is worth pursuing.

First, I didn't reach him, but the next day I did. The guy on the phone was super nice, and when I asked him whether I could come for an internship, he simply said yes.

When I arrived, I realized that I had been right. The plane had never taken off from water because of severe flaws in the hull design, and it had somewhat stopped the whole project and thrown the small company into a depression.

So I set out to understand this problem. I received videos of the test, assembled the different camera angles into one frame and synchronized them in time so that I could view the test from multiple angles at the same time.

I watched the same scene probably more than a hundred times, and every time tried to understand better what was happening. For some reason, each time the pilot tried to accelerate, the wing got sucked into the water and the plane started turning.

Luckily, they had already built a radio-controlled scale model, which made testing a lot faster and easier. So I started quite quickly testing my first hypotheses about what was happening and how it could be changed.

This way I found a solution that was rejected for practical reasons. It was just not feasible to shift the center of gravity in a safe manner during a takeoff run. So I followed a different path towards a solution, using testing to gauge how much I was getting closer to it.

Some ideas seemed very logical at first, like that of creating a hull shape that would “trap” the water under the hull, thus creating lift. Reality showed that the water rather escaped forwards, thus creating more drag than lift.

So my idea of wanting to “force” the water to do what I want got replaced by the idea of trying to imagine what the water would do and then harnessing that movement.

This proved to be a much better mental model, which was exemplified by the model accelerating to a higher velocity under the same amount of thrust. Eventually we had reduced the drag enough for the model to lift all the way out of the water and to start planing, allowing it to accelerate even further.

After this solution was found, we started a new development program for a larger seaplane, which we also tested with a radio-controlled scale model.

I refined my understanding a lot doing more tests and reading into the theory of hydrodynamics and hull design. By now, the company has grown from two to twenty-five employees, and I am still in charge of the hull design.

As the stakes have grown, reading into specific risks related to hull design and how to mitigate them has become more important. I also used CFD simulations to verify improvements that we observed in the model scale. Our latest hull design has a substantially lower drag and spray than conventional hulls, which is quite a success.

In retrospect the recipe seems rather simple:
1. Find a problem that excites you and is really painful for someone else
2. Learn by observation and forming a mental model
3. Test hypotheses and possible solutions – don't be attached to the outcome, only to learning
4. Don't stop. Continue improving and learning more. Learning takes time.