Inside Joss Creswell’s AI-Driven Offshore Sailing

Inside Joss Creswell’s AI-Driven Offshore Sailing

When people ask what offshore racing is, they usually picture big seas, crashing waves, and boats being thrown around in the dark. And yes — it is all of that. But at the level I am racing now, it is also something far less romantic and far more demanding. 

It is a constantly evolving maths problem. 

Every decision offshore is a balance of variables: true wind speed, apparent wind angle, boat speed, sea state, tidal flow, fatigue, and the forecast's reliability. And every six hours, when new weather files are released, the entire board resets. It is like playing chess while someone keeps quietly moving the pieces when you are not looking. 

In 2026, I am racing the Solitaire du Figaro, the toughest single-handed offshore series in the world. There is no co-skipper now. Every decision, every sail change, every routing call, and every mistake is mine alone. That makes the role of data, and how we interpret it, more important than it has ever been.

From Double-Handed to Alone on Deck  

Last season, Next Step Racing competed in the Figaro 3 class as a double-handed team with Charlie Warhurst. That was our first step into this environment: identical boats, world-class sailors, and no hiding places. We learned how brutal the margins are and how much of modern offshore racing is decided not by who sails harder, but by who understands their information better.

That season laid the foundations for what we are now doing with digiLab. 

This year, the training wheels are off. The Solitaire du Figaro is a solo championship, raced across multiple long offshore stages. Competitors navigate, sail, sleep, repair, and think alone. The physical side is tough, but the cognitive load is often the real limiter. When we are exhausted, wet, and cold, our ability to judge risk and uncertainty is what keeps us competitive. 

That is exactly where AI becomes useful. 

The Figaro as a Floating Laboratory  

Every Figaro 3 is identical. Same hull, same sails, same foils. The only advantage we can create is in how we sail and how we make decisions

Modern offshore racing already uses data — weather models, routing software, and performance polars — but most of those tools assume something that is not true: that all data is equally reliable.  

Forecasts are more uncertain in light winds. Wave models struggle in complex coastal areas. Performance polars rarely reflect what actually happens in messy seas. Yet routing software treats all of it as if it were equally precise. 

This is where digiLab comes in.  

For my 2026 Solitaire du Figaro campaign, Next Step Racing and digiLab are building an AI-driven weather-routing and performance-analysis tool that goes a step further. Instead of just producing the fastest route, the system analyses:  

  • historical forecast error  

  • real-world boat performance  

  • wave buoy and satellite data  

  • and model confidence  

The result is not just a line on a chart, but a measure of risk.  

Before a race, I can start to see not only which option looks fastest, but which one is most robust if the forecast is wrong. That matters enormously when I am betting several days of racing on a single weather system.

It turns offshore racing from a guessing game into something closer to informed probability management.  

Caption: This is the polar for the original Figaro; the Figaro 3 can hit 27kts downwind 

Another powerful area where AI is already changing how we race is in refining performance polars; the tables that predict how fast the boat should go at different wind speeds and angles. Traditionally, these are supplied by the sailmakers or the class and treated as fixed, theoretical targets. In reality, no two sailors drive the boat the same way, and no two sea states behave the same. By feeding digiLab’s AI with thousands of miles of real sailing data — including wind, waves, trim settings, and boat response — we can begin to build a living polar that reflects how my Figaro actually performs, not how it is supposed to perform in flat water. That means routing decisions are based on reality rather than theory, allowing me to choose sail plans and courses that match what the boat can truly deliver when it matters most. 

Humans, AI, and the Next Edge    

The Solitaire du Figaro will always be won by sailors, not algorithms. But the sailors who succeed in the next generation of offshore racing will be the ones who understand their data, their uncertainty, and their own limits better than anyone else.  

With digiLab, we are building tools that turn raw numbers into usable insight. With Mustang Survival, I wear equipment that lets me keep functioning even when the ocean is at its worst.  

Between them, they allow me to do what this sport demands in 2026: make better decisions, more consistently, under pressure. And in a fleet where everyone has the same boat, that is where the real advantage now lies. 

Single-handed racing means working in the dark, clipped on, often in heavy seas, while tired and cold, hundreds of miles from land in winter North Atlantic conditions. These are Joss' top Mustang Survival picks that keep him pushing even when conditions get extreme.