For those not familiar with MotoGP, it’s perhaps easiest to think of it as the Formula 1 of the motorcycle world with all the same glamour, technology, superstars, huge budgets and enthusiastic audiences. Technically, it’s at the top of the two-wheeled tree – £1 million-plus bespoke machines racing at over 230mph, supported by a huge team infrastructure that travels to 19 different circuits across the globe every year showcasing 23 riders and 11 teams.
As Davide Tardozzie, team manager of Ducati and the man responsible for making all the human parts work effectively together, told me: “MotoGP is not like the old days of TT racing with riders accompanied by a friend as the mechanic. Today, engineers, particularly electronic engineers, are very, very important.” (In racing circles, data analysts and technicians are classified as engineers.)
Over the last few years, the use of data in motorcycle racing has become vitally important. Telemetry and data analysis allow teams to modify the bike around the individual rider’s style. Mapping and electronic management of the engine enablea riding-by-wire, traction control and variable engine braking which allow skilled riders to extract every ounce of performance from the motorcycle.
A typical MotoGP bike has more than 60 sensors that constantly log different parameters. During free practice on race weekends, each of the four Ducati bikes collects more than 8GB of data on average but, unlike Formula 1 MotoGP, the rules do not allow live data transfer between the bike on the track and the team. This creates a problem as teams need to download, analyse and interpret the data, while also discussing the “feel” of the bike with the riders, and then communicating where improvements might be made, all in just a few minutes.
Working with NetApp, Ducati has built a state-of-the-art, solid-state hybrid cloud data centre that literally travels with the team. It provides very high-speed analysis at the track while simultaneously sending the data back to the engineering centre in Bologna for longer-term strategic analysis. This has given Ducati the edge in dealing with data and, as Tardozzie, said, “I’m proud that Ducati is well known in the paddock for being technically advanced with data.”
John Rollason, director of marketing EMA for NetApp, added: “Secure Data Archive and fast access for analysis is fundamental. It is important to process historical data quickly from last year’s races in order to run tests and take better decisions.”
“Ducati is well known in the paddock for being technically advanced with data.”
Ducati is now utilising machine learning and AI techniques to examine those large volumes of data and speed up its decision making. Gabriele Conti, software and strategies manager for Ducati Corse, commented: “While the high-speed technology has removed many of the limitations on data access, there are usually only five minutes to make decisions about the next session. The tme is never enough, so machine learning is important to understand if you are on the right performance curve.”
The data gathered from test runs is used to create a virtual digital twin of each motorcycle, taking into account the differences caused by each rider’s unique style. This model is then used to help create different base settings from track to track, so the limited testing time that riders have at each race meeting is starting from a base that has already been optimised.
What is particularly striking about Ducati’s approach is the emphasis it places on the relationship between each rider and their engineer, rather than on the observations and insights drawn from data. There is an understanding that riders are complex, “they sense what’s around them – a good engineer has to respect this and use it” when talking to them about data and its interpretation, Tardozzie explained.
Perhaps this is best summed up as, “a rider feels, but an engineer sees”. At the blunt end of racing, softer communication skills are arguably more important at the rider-engineer interface than harder data analysis skills. This is particularly true for a team that needs to inspire its riders to go above and beyond any performance they may already have achieved. What data shows is not necessarily the limit.
Riders are concerned with how their bike “feels”, rather than binary, clinical data terms.
In MotoGP (and other racing), this is often clear from the way that relationships between engineering, data and riders are organised. Ducati, like many teams, only allows one engineer to have a direct dialogue with each rider. This is entirely deliberate as it allows the team to ensure that positive, performance-building messages are constantly supporting the rider in their quest for maximum performance, rather than giving them a literal interpretation of the data, which could appear to be negative (for example, if a test run is below the historical benchmark).
After all, riders are concerned with how their bike “feels” – they interpret their performance as “slip” and “grip”, “speed” and “strength”, rather than in binary, clinical data terms. Engineers who can translate between binary and emotion are therefore very valuable.
The same, of course, is true in business. Data analysts who can talk and empathise with the aims of the organisation will always be valued and driving this kind of empathy into the profession of data will always be important. At the same time, this means that the best engineers/analysts will always be sought after by competitors.
Tardozzie arued that, “when another team wants our engineers, this always has to be a compliment to the skills of Ducati.” Perhaps we should all learn to be so magnanimous! Ultimately, while both commercial brands and racing teams now recognise the value of data, there is something else we all must understand. To move our performance beyond the norm means ensuring we do not simply use data to justify sticking with that norm. Sometimes we can – and should – do better.