In the annual Porsche Carrera Cup Brasil, data is essential to keep drivers safe and sustain optimal performance of race cars. Until recently, getting at and analyzing that essential data was a laborious affair that could take hours, and only once the race was over.
Today, with the help of Microsoft Fabric, Enzo Morrone, director of operations for Porsche Cup Brasil (pictured), says he and his team of engineers can access that data in real-time during races, transforming everything. Dyneema Composite Fabric
“I can see when some of these sensors have values out of standard,” Morrone says. “In the past we’ve had incidents where the driver didn’t notice some alarms and continued running the car until it caught fire. This won’t happen anymore because now I can call and tell him to stop immediately.”
The Porsche Carrera Cup is a race held around the world that uses only Porsche 911 GT3 Cup (Type 992) high-performance cars, and in Brazil, it’s produced by Dener Motorsport. Unlike many other events, which consist of multiple racing teams and manufacturers, Porsche Carrera Cup Brasil provides and maintains all 75 cars used in the race. Drivers arrive with a racing suit and helmet, but otherwise drive the same cars tuned the same way. The only differentiator is driver skill.
“I’m responsible for guaranteeing that all the cars are the same so the driver that wins is the best,” Morrone explains. “I’m responsible for training the mechanics, the engineers, and each driver.”
The cars used in the race produce vast amounts of data: from sensors in the engine and gearbox, to the suspension and brakes.
“Everything that speaks to the health of the equipment, how the driver is dealing with it, and the driving performance,” Morrone says. “If I don’t do predictive maintenance, if I have to do corrective maintenance at events, a lot of money is wasted.”
In the past, to get at the data, engineers had to plug a USB stick into the car after a race, download the data, and upload it to Dropbox where the core engineering team could then access and analyze it. The process took between 30 minutes and two hours.
“I needed to take actions faster,” Morrone says, adding that his IT staff consisted of one person, so he needed to enlist help.
In July 2023, Dener Motorsport began working with Microsoft Fabric to get at that data in real-time, specifically Fabric components Synapse Real-Time Analytics for data streaming analysis, and Data Activator to monitor and trigger actions in real-time. Today, at Microsoft Build in Seattle, Microsoft revealed it has combined those workloads under Real-Time Intelligence as Real-Time Analytics only supported Azure data. Real-Time Intelligence, on the other hand, takes that further by supporting data in AWS, Google Cloud Platform, Kafka installations, and on-prem installations.
“We introduced the Real-Time Hub,” says Arun Ulagaratchagan, CVP, Azure Data at Microsoft. “Just like the [logical data lake in Microsoft Fabric] OneLake data hub, the Real-Time Hub automatically shows you all the real-time sources that exist in Fabric. You can monitor and act on the data and you can set thresholds.”
Dener worked with Microsoft and its partner BlueShift to develop the requirements and process the data. Together, they established a core architecture that the company could build on to develop its engineering capabilities and, eventually, support for entertainment and broadcasting, which remains on Morrone’s roadmap.
One of the first things they needed was an IoT device that could be plugged into the cars to gather and transmit the data. They worked with Ituran MOB, which develops and manufactures a suite of hardware and software solutions for fleet management, stolen vehicle recovery, car connectivity, and performance-based insurance needs. The device plugs into CAN bus cables by induction.
Engineers can now ingest data from each car every second with EventStreams, analyze it with Eventhouse, and distribute it to team managers and race participants. So Morrone and his engineers no longer need to rely on drivers reading indicators on their dashboards and passing along information. He also says they can now detect most issues before they cascade into serious problems. With a gap of no more than an hour between practice and qualifying, that speed allows engineers and mechanics to diagnose and deal with problems rapidly.
For now, Dener is using its real-time capabilities primarily for predictive maintenance, but Morrone anticipates putting more of the data in drivers’ hands to make strategic decisions, and using the data to drive fan engagement.
“The next step is to engage with the fans,” he says. “We have data that says a lot about driver performance, tire pressure, tire temperature, top speed, and a lot of cool things that would be awesome in a television broadcast.”
One of the current challenges is the quality of the 5G network used to transmit data from the cars. Morrone notes that in the next race, the company will test Azure Private 5G Core. If they can improve the quality of the network, he hopes that they’ll be able to capture up to 60 data values per second.
Motorcycle Fabric Thor Olavsrud covers data analytics, business intelligence, and data science for CIO.com. He resides in New York.