Formula 1: How F1 Teams Use Telemetry, Control, and Analytics to Go Faster

Let’s take a step back and go back in time to the first Formula 1 (F1) race in the 1940s. In this race, Kubrick-staring Achille Varzi won in his Alfa Romeo 158. This car was a work of engineering magic for the time, but there’s something else that’s remarkable about it. It was completely mechanical. There were no complicated electronics to control any aspect of it. It stayed like that until the 1980s. For those approximately 40 years, the F1 team engineers had to build and develop their cars without digital aids. That means they had to put the engine on the dynamometer and the chassis in a wind tunnel. The result of this is that those pearly vehicles were prone to breaking down, and engineers had no clue about what influenced their vehicle's performance on any given track. All they had were the fundamental mechanical formulas they learned at engineering school. Without telemetry, control and analytics, engineers were essentially walking in the dark or rules of thumb. They had no idea about what was impacting their team’s performance.  It is crazy how one simple piece of knowledge can change everything. 

Enter Stage Right, Telemetry

Fast-forward to 2022 and the current championship. There are now 300 sensors on every vehicle, and they constantly stream data, including a 1080p video feed on each car back to the garage. You now need a Ph.D. in aerospace engineering to even get near an F1 car. Modern F1 cars even have complicated electrical hybrid powertrains that must work exceptionally well for a team to perform at the top. The teams now spend hundreds of millions of dollars to field two vehicles. Modern telemetry, control and analytics have become instrumental in delivering results for each Formula 1 team. The great news is that the digital revolution with telemetry, control and analytics affected not only Formula 1 racing but also how we engineer modern products.

The Importance of Telemetry, Control and Analytics to F1 Teams

Formula 1 has become dependent on telemetry, control and analytics. It may sound mundane, but the way this information carries F1 team success, you’d think these three are the 1992 all-star dream team. 

Why?

The data streaming off of each car is used for everything from driver improvement to long-term vehicle development. When one driver needs to compare his performance to his teammate, he does this by looking at telemetry data. Additional sensors are attached to the car during test sessions to stream aerodynamic information back to the paddock. This information is useful for determining whether certain wing configurations and aerodynamic changes are working as intended. Telemetry, control and analytics are also essential pieces of the puzzle regarding engine performance and vehicle reliability. 

Reliability is now a more significant part of Formula 1 because each power unit has to last much longer than in earlier years. The power unit is also more complicated because you now have a thermal and electric energy recovery system in addition to the batteries and internal combustion engine. The importance of telemetry, control and analytics to F1 teams means that the system needs to work reliably at every racetrack. One failure can derail a team's race, potentially causing them millions of dollars in the lost prize money from finishing lower in the race and season. This telemetry data is used to develop strategies and potentially spot ways to optimize equipment as the driver goes through the race. Before the race, this data is used for testing and setting up the vehicle to perform optimally at the track.

The Architecture That Makes This Process Work

With 300 sensors on each car, recording data and getting it back to the team in seconds is an immense challenge. The system design must be robust and purpose-built for Formula 1 teams. Formula 1 uses a customized mesh wireless network system based on WiMax 802.16 at each racetrack. The sensors record data, which is then temporarily stored in the ECU. This ECU is a custom-built model from McLaren, and all teams use it. That sensor data then travels wirelessly to a centralized location managed by Formula 1. It then securely transmits that data to the relevant team. In each part of the process, everything has to be built to be as robust as possible, as one single problem could derail the entire system. 

The Formula 1 teams use ATLAS to view and analyze sensor data. The ATLAS viewing and analytics suite is also built by McLaren. Teams also send the data from the paddock back to their factories, where hundreds of engineers are watching everything in real-time. Teams also have an "umbilical cord" they can attach to the car to extract data much faster. That is one of the tasks that happen when a driver returns to the garage. AI and machine learning have become indispensable tools along with cloud computing solutions in this process. These cloud solutions allow F1 teams to get insights from their telemetry data quickly. Cloud solutions also give teams access to essentially unlimited compute power without having to manage a hardware cluster themselves. A Michael Jordan, if you will. This system provides the reliable data processing that teams need. 

Challenges with Telemetry, Control and Analytics for F1 Teams

The system design challenges stemming from F1's telemetry, control and analytics architecture are immense. F1 cars are notoriously fickle, which makes it even more challenging for teams to make this system architecture work. Each Formula 1 vehicle has to be made to be as fast as possible, which means that teams prioritize weight savings over robustness when designing parts. Aerodynamics is also the main determining factor in how well a vehicle performs. That means the packaging of components inside the car is also critical. The hybrid system generates much more heat than previous F1 car designs. That system has a significant impact on how and where sensors can go. 

Furthermore, you also have the issue of the ECU failing, leading to data loss if it becomes unrecoverable by the team. There are also security concerns from transmitting data wirelessly at each racetrack. On top of that, you also have problems ingesting and utilizing that data to provide value to each team. Each car sends about 1.1 million data points that can be analyzed to give teams a competitive edge. If teams aren't using that data effectively, they are essentially putting themselves at a significant disadvantage.

How Cadence Uses Inspiration from F1 to Transform the SoC Design Process

The SoC design process is a lot like F1. Chip design companies are in an arms race where they try to one-up each other in every design iteration to ship the highest performing product to consumers. Cadence is a major enabler of SoC design companies, which is why we understand how important it is to leverage telemetry, control and analytics to develop the best SoC design tools for the marketplace. 

Like F1, we leverage machine learning and artificial intelligence algorithms to make our SoC design products faster and more reliable. Many tasks in the SoC design process can be optimized or automated using these machine learning algorithms. Tasks such as verification and physical design are dramatically improved when using ML and AI. These technologies are being integrated into our core technology to help you bring your products to market faster.

Learn how Cadence can help you create better products much faster today!