As of late, Andrea Dovizioso has been trying to find his pace with the new 2017 MotoGP bike while on the other hand, Ex-Yamaha Rider Jorge Lorenzo is still getting himself acquainted with the Italian racing machine.
Between the last testing that took place at Phillip Island in Australia and the upcoming one that is scheduled to take place at Losail, Qatar on March 10-12, the team of engineers working behind the two riders are the ones who study and analyse every tiny bit of data collected from these test sessions. To harness and make the best sense out of the information gathered, Ducati as of late, has started using machine learning technologies.
Data processing isn’t what it used to be in the early days of computer-programmed racing machines. Today, data comes in huge lumps after each track outing which needs to be studied to understand how the motorcycle’s suspension, brake, engine and fueling systems are behaving with respect to a lot more parameters such as track surface and temperature.
“There are 18 MotoGP tracks, and to make sure our bikes perform to their limit, we need to test as many configurations and scenarios as possible,” said Luigi Dall’Igna, Ducati Corse General Manager. “So far, we’ve seen excellent results in the lab with the Accenture solution. The ability to use existing and new testing data will help us choose the optimal configuration for our bikes. This innovative tool will make our testing a more intelligent process, helping us get the best performance from our bikes, whatever the weather or the track.”
Just to bring your head around the amount of data produced on today’s MotoGP bikes, Accenture, the official digital partners of Ducati say that they use more than 100 IoT (Internet of Things) sensors on every motorcycle. Each of these sensors produces data regarding the motorcycle’s performance under a variety of testing conditions. This process is carried through 18 MotoGP tracks every year. That’s a serious amount of data beyond the comprehension of any single human to handle manually.
So it’s certain that a manual procedure to extract the insightful data won’t be feasible or economical, and could still lead to the eventual human error. For these MotoGP machines that are tuned to such levels of precision, even the smallest of errors would result in loss of precious performance. So the best and quite frankly only solution henceforth is machine learning.
“With this solution, Ducati Corse can operate an intelligent testing program to help deliver better race results,” said Marco Vernocchi, EALA Lead of Accenture Analytics, part of Accenture Digital. “By simulating and monitoring a motorbike’s performance under a vast array of track and weather conditions, we’ve been able to apply machine learning integrated with IoT sensor data, to help minimise the time, expense and effort of traditional on-track testing. Ultimately, we hope this innovative solution will help the Ducati Team stay ahead in every race they compete in.”
According to Accenture, more than 4,000 sectors of race tracks under 30 different scenarios have been analysed till date via their intuitive machine learning tools. These tools when fed with the massive amounts of data produced by the 100-or-so onboard sensors present test engineers with a set of the best possible outcomes that could be applied to Ducati’s MotoGP bike for that particular track and condition.