Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage News Ferrari Chooses AWS Machine Learning for Racing and Road Operations

Ferrari Chooses AWS Machine Learning for Racing and Road Operations

This item in japanese

Ferrari will use AWS machine learning, analytics, and compute capabilities across the organization, including the Formula One team Scuderia Ferrari.

The Italian automotive company has announced the agreement, which is both technological and commercial. Ferrari will use AWS services to simplify the design and testing of its cars, offering customers new driving experiences. Additionally, Scuderia Ferrari plans to use AWS to launch a digital fan-engagement platform via its mobile app. 

Ferrari will use AWS's global infrastructure, including the AWS Europe (Milan) region, and breadth and depth of services to streamline design and testing of its cars, giving customers better driving experiences.

Ferrari will leverage Amazon Elastic Compute Cloud (Amazon EC2), with access to specialized instance types for efficient high-performance computing (HPC) and will use AWS Graviton2-based instances. 

Ferrari will build a data lake with Amazon Simple Storage Service (Amazon S3) and use AWS Lake Formation to quickly and securely gather, catalog, and clean hundreds of petabytes of data and will also leverage AWS to make it easier for current and prospective customers to build, purchase, and maintain their cars. 

Using Amazon Elastic Kubernetes Service (Amazon EKS) and Amazon DynamoDB (AWS’s fully managed key-value database), Ferrari will be able to quickly create, deploy, and scale improved digital experiences such as the Ferrari Car Configurator.

By using notebooks in Amazon SageMaker, the Ferrari Formula One team will be able to analyze and visualize large quantities of timing, tyre, and weather data uploaded to Amazon S3 to understand how the race looks from an algorithm’s point of view. The team can analyze strategies and outcomes from past races and debate which historical data points would be most valuable to extract and ingest into ML models, as well as which data points might be most valuable to include in real time during a race. 

Finally, machine-learning models built on Amazon SageMaker can provide insights into developing driver battles during the race that are not always obvious to the audience, like striking distance and predicted overtake difficulty. Having cleaned all relevant data from various sources, the team quickly set business and technical key performance indicators (KPIs), technical requirements, and output formats to be used in validation code that allow for quick experimentation with feature engineering and various algorithms to optimize for the prediction error during every race.

The sports industry's use of AI technologies is driving shifts in real-time analysis, automation, prediction, and dynamic information.

Real-time analog data collected and analyzed by hand means that live commentators provide color, and nothing else. Now fans, on-air talent, and teams benefit from automated data collection through sensors and cameras, and high-performance computing means that information and analytics created with machine-learning models, such as the impact of an F1 car that chooses to pit, can be processed in real time.

When it comes to automating routine, unreliable and time-consuming tasks, it frees up valuable resources. From machine-learning models that forecast ticket sales or predict the likelihood of capture, to NASCAR that uses artificial-intelligence services to automatically tag media using object detection and speech-to-text translation, artificial intelligence enables humans to do what they do best: more creative and strategic work.

From reactive to predictive, coaches and teams are forced to constantly react and change strategy, or anticipate a play or action based on pure intuition; they now have access to a treasure trove of predictive data during a match, game or race, giving them the ability to make proactive, informed decisions in real time during a game, like an NFL coach anticipating a play just before time runs out.

Any fan of sports from static to dynamic will tell you that their sport is as much a mental game as it is a physical game. By giving fans access to data and insights through visually rich on-screen graphics and interactive second-screen experiences, sports organizations and broadcasters can open the curtain on the nuances of decision-making, enriching experiences. from fans both inside and outside the stadium, like Six Nations Rugby displaying real-time information during a ruck.

Rate this Article