Video: Formula E car charges battery during mountain downhill drive to Monaco

Formula E has published a video of a PR stunt in Monaco. Former racing driver James Rossiter steered the electric series' so-called GenBeta car down a mountain passage to Monaco - with only one per cent energy. Rossiter's task: to recover enough energy for one lap of the Grand Prix circuit through recuperation alone.

In collaboration with Google Cloud, Formula E had its development racing car roll 1,000 metres down a French Alpine road. The vehicle collected enough energy to complete a full lap of the racetrack using braking energy recovery (recuperation) alone.

The project, filmed and produced by Red Bull Media House in partnership with Google Cloud marketing, not only demonstrated the car's performance on the mountain track from the summit of the Col de Braus, but above all the recuperation performance of the GenBeta car.

Background: Although the 22 Formula E drivers start with full batteries, they have to generate at least 40% of the energy they need during each race through recuperation. This makes the cars the most efficient and sustainable racing cars ever. In race mode, the vehicle can recover 600 kW of energy (250 kW via the front and 350 kW via the rear drive unit).

Collaboration with Google Cloud

For the realisation of "Mountain Recharge", Formula E cooperated with Google Cloud, its official partner for cloud technology and security. Support was also provided by GenBeta partners ABB, Hankook and SABIC.

With the help of Google Cloud's AI-powered solutions, in particular the Gemini API via Google AI Studio, Formula E and Google Cloud first tested the feasibility: Can the GenBeta race car generate enough energy for a full 3.337 km lap of the Circuit de Monaco by recuperation when rolling downhill?

The following technology was used:

• Google AI Studio: The Gemini API was used via Google AI Studio to analyse the complex variables of the descent. The AI model helped to find optimal braking points, calculate the influence of speed, weight and gravity and optimise the riding style for maximum energy recovery.
• BigQuery: Google Cloud's serverless data and AI platform collected, stored and analysed real-time telemetry data from the vehicle during descent for review and analysis.
• Firebase: A dashboard application was quickly created using the Firebase application development platform. This allowed race engineers to visualise the vehicle data from BigQuery in real time on their Chromebooks and Pixel devices.
• NotebookLM: Google's AI-powered research and writing assistant NotebookLM helped to bundle technical data, logistics plans and engineering data during the project and facilitate collaboration between the teams.

By utilising the potential energy of gravity alone, the car was able to recover between 1.6 and 2.0 kWh of energy - enough for the entire 3.337 km race track. That's enough energy to fully charge almost 60 Google Pixel 9 Pro XL smartphones, just through recuperation.

"With the GenBeta project, we want to push boundaries"

"This is more than just a spectacular racing story - it's an example of how high-efficiency recuperation and cloud-based AI can revolutionise the way we look at mobility, energy optimisation and sustainability," said Alex Aidan, VP Marketing at Formula E.

"At Formula E and with the GenBeta project, we want to push boundaries and do things that others have never even dreamed of. The rapid development of our partnership with Google Cloud and our long-standing relationship with ABB demonstrates not only the potential of technology to transform racing, but also how global brands are bringing their stories and products to life through impactful collaboration."

Guillaume Roques, Senior Director, EMEA Marketing, Google Cloud, adds: "Google Cloud is happy to help partners overcome unique challenges using data and AI. The 'Mountain Recharge' project is an impressive demonstration of how AI can tackle complex, real-world problems. With our technologies, we were able to model the complicated physics of the descent and accurately calculate the potential for energy recovery."