How artificial intelligence and virtual reality are helping Ducati to dominate MotoGP

MotoGP

There are all the usual reasons we talk about when discussing Ducati’s MotoGP success, but underpinning it all is something that no one really talks about: computer modelling, virtual reality and artificial intelligence…

Cristhian Pupulin and Jack Miller debrief with telemetry in Ducati MotoGP garage

Tablets, laptops and computer servers are now as much a part of MotoGP as spanners and torque wrenches. Engineer Cristhian Pupulin (left) and Jack Miller both take their Ducati knowhow to KTM this year

Ducati

display_61b60c3443

“I can’t believe how far Ducati are ahead,” one top MotoGP engineer told me shortly before I travelled to yesterday’s 2023 Ducati launch in upmarket Italian ski resort Madonna di Campiglio. “The power they can put down out of corners is insane – if they had a Marc Márquez or a Fabio Quartararo on that bike we wouldn’t see which way they went.”

That’s how it is in MotoGP right now.

We all know Ducati has more horsepower than everyone else – when the company entered MotoGP in 2003 it led the top-speed race at 206mph (332.4km/h), now it leads the way at 225.7mph (363.3km/h).

Having eight bikes is such an advantage, because the more data you get the more accurate the modelling

We also know that Pecco Bagnaia’s 2022 championship came from a chassis that now turns corners as well as the engine goes down straights, thanks to Formula 1-inspired diffusers that create grip out of the air, amongst other upgrades.

Something else we all know is that Ducati has eight bikes on the grid, so it gets huge amounts of data to fine-tune its Desmosedicis. Eight data streams also help any Ducati rider struggling with a particular corner or section, because he can look at seven other ways of doing that corner or section to learn and adapt his technique.

But perhaps Ducati’s biggest advantage of them all – because it underpins all the above – is something very new that most people don’t know or talk about: computer modelling and simulation.

Over the last few years Ducati has invested massively in this technology, which is improving at a dizzyingly fast rate, with AI (artificial intelligence), VR (virtual reality) and augmented reality now an important part of what Ducati does, both at the racetrack and back home in Borgo Panigale.

Pecco Bagnaia and Bagnaia and Cristian Gabarrini examine MotoGP telemetry

Bagnaia and Cristian Gabarrini examining data last year. Can you guess the track?!

Ducati

“This technology has increased a lot, a lot in the last couple of years,” says Ducati race boss Gigi Dall’Igna. “We also did many things like this in the past but the results weren’t good, because you need to build up the data. Now AI is really important for us to achieve our results.”

Of course, other manufacturers are also working in these areas but Ducati is currently way ahead.

The company now has more engineers in the MotoGP paddock than anyone else, with many of its red-shirted hordes not spinning spanners but tapping away at computers – interpreting the data, then modelling and simulating performance in every scenario, because this is the new way to get ahead, as pioneered in Formula 1 cars, aerospace and many other industries.

Related article

AI and VR are already huge in F1, with top teams using super-powerful computers and software to run hundreds of thousands of virtual laps (yes, hundreds of thousands) before each GP weekend.

Ducati does the same thing, but on a smaller scale, because MotoGP budgets are many times smaller than F1 budgets, and because modelling a motorcycle is much more difficult than modelling a car, because motorcycle dynamics are way more complex than car dynamics.

“Because of what the bike does and the fact that the rider is always changing position makes it much more complicated to model a bike than a car,” affirms Dall’Igna.

This is another reason why having eight bikes is such an advantage, because the more data you get the more accurate the modelling.

Ducati MotoGP engineer looks at tyres and tyre data on a tablet computer

Tyre data is arguably more important that anything else at the moment because Michelin’s slicks require deep understanding to work at their best

Ducati

Ducati uses its vast data haul to run individual laps and race distances, via AI, VR and augmented reality. This allows its engineers to visualise how a tyre will perform over a single qualifying lap and how it will degrade during a race, what effect a geometry change, suspension adjustment or mapping tweak will have at certain stages of a race and so on. Thus engineers can find the right answer to improve real-world performance faster and more accurately, so they pretty much know how the motorcycle is going to behave before it leaves pitlane.

The advantages of modelling and simulation grow year by year, not only because the technology gets better, but because there’s less and less testing and practice time allowed, so virtual testing, practising and racing has a greater effect on results.

Related article

MotoGP manufacturers have built virtual 3D models and maps of race tracks for more than a decade, using LiDAR scanners, which are walked around tracks on trollies to record everything, from overall layout, corner angles, camber, environment and so on. But effectively correlating that to a computer-created model of their motorcycles is much more recent.

This is why Ducati’s title sponsorship with Lenovo isn’t merely a tool to flog more laptops, it’s a technical partnership that has Ducati using central high-powered servers via virtual desktops in the team garage, the team truck and in the factory race department, where, of course, modelling and simulation has a huge effect on short- and long-term machine development.

And this is one reason why some factories are poaching engineers from Ducati, most importantly Enea Bastianini’s and Andrea Dovizioso’s former crew chief Alberto Giribuola, who now oversees the factory KTM garage. Giribuola is a graduate of mechatronics, which mixes mechanical engineering, electronic engineering and robotics.