America’s Cup World Series: How Sir Ben Ainslie’s Land Rover team uses artificial intelligence technology to seek an edge
Sir Ben Ainslie and his Land Rover team set sail off the coast of Portsmouth this weekend for the America’s Cup World Series’ sixth leg, aiming for a repeat of their triumph on Southsea Common.
But it won’t just be the five-time Olympic medallist and his crew’s grappling to steer control of their 44ft foiling catamaran while flying at high speeds of 37-38 knots who will be responsible should they do so.
With over £100m being ploughed into the Land Rover Ben Ainslie Racing (BAR) team from wealthy backers including Sir Charles Dunstone and Sir Keith Mills with the singular aim of bringing sport’s oldest trophy to the UK for the first time, no potential avenue for a competitive advantage has been ignored with even cutting-edge artificial intelligence technology now being deployed.
Described as the Formula 1 of the seas, the team of designers and engineers behind the scenes working on the boat that raced by Sir Ben and co are equally if not more important than their equivalents in the famous road race.
Even Ainslie himself has acknowledged the outcome of a race is “50-50” between “design and actual racing skill”.
No wonder then that Land Rover was prepared to pull 27-year-old MIT whizz kid Mauricio Munoz from the company’s advanced engineering team and into the Land Rover BAR set-up to use artificial intelligence technology in the design process.
“I’m on the bleeding edge of trying to find ways and methods that the performance team can analyse and evaluate the performance of the boat,” Munoz told City A.M.
The Land Rover BAR team receive over 16 GB of uncompressed data every day — the equivalent of an iPhone’s entire memory — picked up by sensors on the boats measuring more than 300 variables.
Insights gleaned from this data have the ability to make the difference in a race if it can help the design team enable a stable flight on the foils and keep optimum distance between the boat and water for longer.
But making sense of such an overwhelming amount of information is a tall task even for MIT grads. That’s where artificial intelligence technology comes in.
Think moneyball for boats rather than creating a self-sailing catarmaran at this stage.
“It’s not artificial intelligence per se but the techniques that underly artificial intelligence, what I call machine learning,” Munoz explains.
“It lets you pick up the patterns and look at data in a really interesting way. Artificial intelligence would take it a step further by telling you 'I know these patterns, I know that this is how the world works and what’s the best decisions to take at any given point'.
“Here it’s more about trying to figure out what those patterns are and letting the sailing team figure out the best use of it they can.”
If everything goes to plan, the machine learning technology could be the marginal difference between triumph and disappointment when Land Rover BAR goes for the ultimate prize, the 35th America’s Cup, in Bermuda next year.
For now, it’s about putting themselves in the best possible position for victory in that race — and victory in Portsmouth would help.
The winner of the World Series is awarded two points for the qualifier round of the America’s Cup proper, while the runner-up — the sport currently occupied by Land Rover BAR — receives one.
With just two regattas to follow in Toulon and Fukuoka after this weekend, the stakes are increasing, and every insight made from machines and Munoz carries a potentially greater influence.
“What I do is another cog, another wheel in the big machine that is design and performance analysis. It has the potential to have a really, really big impact,” he says.
“You can’t really overstate how valuable that is. Sailing is such a dynamic and challenging environment that it’s unlike Formula 1 in the sense that if you put the same boat on the same course on two different days you’re going to get two different performances.
“With Formula 1 usually you put the same car, same driver down the same track on two different days with more or less the same conditions you’re going to get approximately the same results whereas here you’re going to get something totally different.
“You’re playing with a problem that’s just on a whole different dimension.”
Need to know
[infographic id="849"]