This week marks a milestone for me — and possibly a first in the history of iceboating.
I have just finished building a runner plank designed entirely by artificial intelligence.
Over the past days, I trained an AI model by feeding it detailed information about more than 100 runner planks built over the last decade — including materials used, layup schedules, structural failures, stiffness measurements, field results, and performance notes. Based on this dataset, ChatGPT proposed its own optimized layup concept… and the design was so interesting that I decided to build it.
The plank is now finished and curing. It will touch the ice later this year, most likely in Sweden.
What convinced me to actually build this experimental component was something remarkable:
ChatGPT calculated precise static deflection of my plank without knowing my design targets — only by analyzing the materials, geometry, and fiber orientations. Its estimate was nearly perfect.
Even more importantly, AI allowed me to perform structural calculations that I had never been able to do before — complex stiffness modeling, stress analysis, core compression predictions, dynamic load simulations. These are things I used to approximate based on intuition and experience, but now I could verify them with real numerical feedback.
Whether AI can help us build better iceboat components… we’ll find out soon. For now, I’m genuinely excited to see if this pioneering plank will prove itself on the ice.
Stay tuned — the experiment may only beginning.
