Formula 1: Oracle Red Bull Executive Declares Sport the Ultimate AI Proving Ground

Formula One is poised to significantly expand its use of artificial intelligence, with the sport serving as the premier platform for showcasing this advanced technology, according to a key executive at Oracle Red Bull Racing. "Data permeates every aspect of our operations," stated Jack Harington, the F1 team's technical partnerships lead, in a conversation with The National at Oracle AI World in Las Vegas. He noted that Formula One has seen immense technological advancement in recent years, with the integration of new technologies accelerating dramatically. "I believe it's the ultimate showcase for technology. We are deeply dependent on our technology partners to supply us with the tools necessary for operating our cars and managing our business. It's a fertile ground for testing new technologies." Formula One has been at the forefront of AI adoption, with each car equipped with 300 to 600 sensors that monitor critical variables like engine temperature, tire pressure, brake wear, and aerodynamic forces, as reported by IMD. Nearly every Formula One team collaborates with major technology firms, including Microsoft, Oracle, AWS, Google, Dell Technology, AMD, and Palantir. These teams leverage cloud infrastructure and real-time data analytics to conduct billions of simulations, expedite car design processes, optimize race strategies, and gain a competitive advantage, according to a report from IMD last year. Mr. Harington refrained from specifying the exact number of sensors on the cars, citing the teams' secrecy surrounding their AI strategies. "These cars are data-driven powerhouses, and we must continuously monitor every component. I suspect that if left to the engineers, we would have thousands of sensors, but each sensor adds weight, which is a critical factor in Formula One. Therefore, we must minimize the number of sensors to ensure competitiveness," he explained. "We likely extract approximately two terabytes of data from each car during every track session. This data contains millions of points that we must analyze to confirm our car's performance aligns with our expectations and to maximize its efficiency over a lap or an entire race." In preparation for a race weekend, the F1 team executes eight billion Monte Carlo race simulations to determine the most efficient path from the start to the finish line. One practical application of this data involves strategy simulations, Mr. Harington elaborated. "In F1, we deal with two types of data: 'known data,' such as tire degradation patterns or tire lifespan, and 'unknown data,' which includes unpredictable events like yellow flags, red flags, unexpected overtakes, or safety car deployments." The team employs Monte Carlo simulations, a computational method utilizing repeated random sampling to predict the likelihood of various outcomes, prior to races. Before a race weekend commences, the team runs eight billion Monte Carlo race simulations to map out the fastest route from start to finish. During practice sessions, all relevant data is collected. "We then incorporate this data into our simulations and run an additional eight billion Monte Carlo simulations. If an unforeseen event occurs during the race, we can update these models in real-time," Mr. Harington stated. As part of their partnership with Oracle, the team has migrated these simulations to the Oracle Cloud Infrastructure. "This offers a significant advantage to our team, as the scalability of simulations enabled by Oracle Cloud allows us to conduct extensive 'war games.' We not only simulate the fastest race strategy but also predict how rival teams might respond to our decisions," he concluded.