Real time analytics – This is how Red Bull Racing does it
Real-time analytics is a must for Red Bull Racing to be able to optimise their cars for speed. ”Fast” – that doesn’t sound too hard. Just drive fast enough, and you’ll win – even in major races like Formula 1. Well, no – it’s not as simple as that. For optimal speed, you need a complex IT configuration.
Matt Cadieux, the IT manager of Red Bull Racing, says that his team makes their cars perform at maximum level by using the Internet of Things (IoT).
Still, of course, it all begins with the car itself. F1 cars are low-flying, lightweight aircraft with an engine capacity of approximately 600 horsepower. The cars have a fibreglass body and weigh only 680 kilograms (1,500 pounds). They’ll go from zero till 100 km/h in about three seconds, and they will protect the driver even if something goes wrong and the car crashes.
A car with 200 sensors
The conflicting requirements that the car should get the best results possible while at the same time comply with the regulations is a challenge. Only a certain amount of horsepower and downforce are permitted. Fuel levels must also be within certain limits. Red Bull is constantly redesigning and redeveloping its F1 cars to achieve the greatest effect possible right up to the qualifying rounds for every major race. The engineers can make changes up till then.
A large part of the work at Cadieux’s department is about making sure that the engineers have access to all information they need to make these decisions. A complex, fully equipped IoT system connected to a fast and flexible infrastructure is a must.
No changes in car configuration permitted
During the tests, the cars have around 200 sensors, but in the actual race, the number is somewhat reduced to make the car lighter. Everything is measured. From the physical forces that the car needs to handle to engine temperature, load, and aerodynamic information.
”That way, we can see the inside of the car without having to open it. For example, we can see what physical forces the car needs to withstand and what parts get overheated. We can also see what happens aerodynamically if our predictions are correct, and if what happens in the wind tunnel reflects the reality on the racetrack,” says Cadieux.
The system encrypts all data and transfers them wirelessly to the Red Bull team at the track. The team can find out a lot from real-time information even before the race is finished. They can, for example, see if there are performance problems and tell the driver that he needs to compensate for that. Once they’re qualified for the race, the team is not allowed to change the configuration of the car. Any changes that may be called for must be done before that.
Mobile data centre by the race track
The team does not need to do very many complex calculations by the race track, even though there is a mobile data centre next to the track. It has an MPLS connection and the XenDesktop virtualization system from Citrix. All complex analysis is done at the Red Bull data centre in the UK, where in the past year the hyperconverged infrastructure SimpliVity from HPE has been used.
Neil Bailey, head of IT infrastructure at Red Bull Racing, stresses the importance of infrastructure:
”The system affects how fast we can react and thus how fast we can make changes to improve the performance of the car. The cars aren’t just delicate machines driven by top athletes like Max Verstappen, they’re also high-tech platforms which, during every racing weekend, generate around 400 GB of data, which the team needs to analyze as soon as possible. Based on these real-time analytics, critical decisions need to be made very quickly, for example, whether a depot step is needed.”
30,000 changes each year
The stable used to manage this from the UK, using different traditional virtualized servers and infrastructure of virtual computers. In all-around 500 virtual machines on various types of hardware. This made for a fragmented and divided system where around 50 TB of data were circulating. This, in turn, leads to delays for the software engineers at the virtual computers, but Red Bull Racing relies on a fast and responsive infrastructure for its F1 calculations.
”All our cars are prototypes that are being continuously developed and subject to around 30,000 changes each year. It is extremely important that we have an infrastructure that is very flexible to meet the needs of our engineers and the ever-changing demands that they are facing,” says Matt Cadieux.
”Because every new year comes with new competition challenges, technical improvements and changes in the F1 rules. So if you want to be successful in this sport, you need the right IT infrastructure.”
Real-time analytics – Every second count
The new infrastructure offers obvious advantages compared to earlier systems, not least because everything is perfectly coordinated. The team receives real-time data from the car during racing days and processes them for analysis. What used to take nine minutes now takes two.
”This is an amazing improvement which means that we get faster and better answers on the track, where every second count,” says Cadieux.
This shows how a simplified and optimized infrastructure directly affects what happens on the track. Also, virtual computers have become faster, and backup and recovery in the Citrix VDI environment have been improved. When the team for research and development would discover an error in the system, they used to have to reset their virtual workstations by applying complex processes. With the new system, all that is needed is three clicks, and then reset itself takes about a minute, something that Bailey is very pleased with.
”It used to take more than an hour and required several complicated measures.”
For Red Bull Racing, it is extremely important to be able to analyze large volumes of data rapidly to be able to apply the data in real-time while the race is running. This requires a very flexible infrastructure, which Red Bull Racing now has acquired, by using the HPE hyperconverged infrastructure SimpliVity.