GCP – Formula E’s AI equation: New Driver Agent boosts next generation of racers

By its very nature, Formula E has always been driven by innovation, bringing the next-generation of vehicle power to the racing circuit. An equally important mission has been democratizing access to these new technologies. 

Formula E has been working with Google Cloud to set the technical foundations for many of its innovations. Now, that collaboration extends to helping level the field, bring more excitement to the track, and, above all, inspiring and empowering the next generation of racing talent. 

Recognizing that not all drivers have access to the same level of data and analysis, Google Cloud and Formula E are leveraging AI to help bridge this gap. They’re also partnering with organizations like More than Equal to provide cutting-edge AI tools and resources to support the development of diverse talent in racing.

The AI-Powered Driver Agent

At the heart of this initiative is the “Driver Agent,” an AI tool powered by Google’s Vertex AI platform and Gemini, Google’s state-of-the-art AI foundation models. The Driver Agent is designed to analyze extensive multimodal data generated during racing and provide actionable insights to drivers. This data is in a number of formats such as text, tabular data, and images of telemetry graphs and heatmaps, and includes key performance indicators such as:

• Lap times
• Speed
• Braking
• Acceleration
• G-force
• Downforce
• Latitude & longitude
• Steering

Here’s a look at how the Driver Agent works:

• The racing simulator creates detailed logs of the car during the race, which are sent over to Google Cloud, where they land in Google Cloud Storage

• An event based trigger via Pub/Sub triggers a Cloud Run service to start, where pre-processing of the race logs are performed. This includes:

• Identifying the fastest lap from the race

• Creating interactive lap heatmaps for display in the UI, as well as static images of the race heatmap to share with Gemini

• Creating certain BigQuery Tables and Views out of the log data, ready to be queried and shared with Gemini

• Joining Table data from the incoming race with the comparison data of a race by a Professional Formula E driver, allowing Gemini to compare and critique vs this professional lap

• Creating a series of interactive (for the UI) and static (for Gemini) graphs showing particular views on acceleration, braking and lateral g-force

• The Frontend, driving the UI

• The Backend, an API server that orchestrates retrieval of the pre-processed data sources and the requests and responses sent to Gemini.

Democratizing Data for All

The Driver Agent plays a crucial role in democratizing access to data-led analysis and coaching. It connects to a Formula E driving simulator and provides in-depth, immediate coaching through both text and audio. This is achieved by analyzing low-level telemetry data from a user’s lap (including latitude, longitude, speed, brake %, g-force, downforce, etc.) and comparing it with data from a professional driver on the same track.

Supporting More than Equal

Google Cloud is actively supporting More than Equal’s mission of identifying and developing elite female racing talent. This support includes:

Providing access to advanced driving simulators at Manchester Metropolitan University Institute of Sport and Centre of Excellence for Women in Sport, the training facility for More than Equal development drivers.

Providing the Driver Agent technology to five developing drivers within More than Equal’s Driver Development Programme, to aid in their training.

This tangible commitment from Google Cloud provides the More than Equal Driver Development Programme with valuable access to AI Agent technology.

Extending the Impact with Formula E

In addition to supporting More than Equal, Google Cloud collaborated with Formula E to further develop and refine the Driver Agent. Formula E, at the pinnacle of electric motorsport, provides a valuable platform for showcasing and advancing AI applications in racing. The Driver Agent is being used to analyze data from Formula E races and provide insights to drivers, helping them to:

• Improve lap times

• Optimize braking and acceleration

• Generate more downforce

• Avoid crashes

The Driver Agent has demonstrated significant potential, with tests showing it can help drivers improve their lap times. Furthermore, the collaboration between Google Cloud and Formula E is helping to level the playing field in the sport by making sophisticated AI tools accessible to a wider range of teams.

“Formula E has always been a platform for innovation, and we are thrilled to partner with Google Cloud to push the boundaries of what’s possible when you combine world-class technology with world-class motorsport.” said Beth Paretta, VP Sporting at Formula E. “By developing and offering these cutting-edge tools, we are helping to create a future where racing talent is determined by skill, not resources, enabling a more diverse pool of drivers and especially women, rise to the very top of our sport.”

Conclusion

Formula E and Google Cloud are dedicated to using its AI technology to positively impact the world. By focusing on initiatives that promote inclusivity and provide access to advanced tools. Helping to shape a future where talent is not limited by opportunity.