Formula E is expanding its Google Cloud partnership beyond sponsorship into the machinery of running a global racing championship. Gemini models, digital twins, real-time analysis, and workplace tools will support event logistics, operational planning, performance, and broadcasts.
The central challenge is unusual but instructive. Formula E moves cars, equipment, temporary infrastructure, and people between race locations while maintaining its status as a sport certified net zero carbon since inception. Improving a lap time and reducing emissions both depend on turning complex physical systems into decisions.
AI becomes valuable when it reduces unnecessary movement before the first truck leaves.
Digital Twins Make Physical Logistics Testable
Race and event digital twins allow Formula E to model site builds virtually. Teams can examine layouts, infrastructure requirements, and operational alternatives without relying on as many physical reconnaissance visits or transporting equipment simply to discover what works.
Those avoided journeys and adjustments matter for Scope 3 emissions, which sit across a supply chain rather than inside one office or vehicle. Virtual planning can expose a lower-carbon option while changes are still inexpensive, then provide a shared reference for distributed teams.
The approach also improves resilience. A digital representation can help planners compare scenarios when a location, schedule, or requirement changes. The strongest benefit is not the visual twin itself, but the ability to connect its recommendations with procurement, transport, staffing, and event decisions.
Formula E is also deploying Google Workspace with Gemini to reduce administrative delay. These everyday tools are less dramatic than a racing simulation, yet faster coordination can be important in an organisation that repeatedly assembles and dismantles complex venues.
Racing Data Links Efficiency With Performance
The earlier Mountain Recharge project demonstrates how the partnership applies AI to a physical problem. Engineers used Google AI Studio and Gemini to map a mountain descent for the GENBETA car, analysing topography, friction, energy use, and braking zones.
The aim was to calculate how regenerative braking could harvest enough energy during the descent to support a subsequent full lap of Monaco. This is a concrete optimisation problem: the model’s recommendation must survive gravity, tyres, weather, driver execution, and energy constraints.
Such projects make motorsport a useful laboratory. Milliseconds and energy flows are measurable, and the consequences of a poor assumption appear quickly. Techniques developed under racing pressure may inform wider work involving electric mobility, simulation, and efficient physical operations.
Formula E should still separate a showcase from a repeatable capability. The relevant question is whether the same data practices and modelling tools reduce emissions or improve performance across seasons, not whether one demonstration attracts attention.
Real-Time Insight Is Also a Fan Product
Formula E has embedded a Strategy Agent in live broadcasts. It processes race data to explain strategy and driver performance, helping viewers understand decisions that unfold too quickly or involve too many variables for a conventional commentary alone.
Millions of viewers have used these insights, according to the source. The application resembles enterprise observability: a high-volume technical stream becomes an understandable narrative for people who need context rather than raw telemetry.
The commercial value is distinct from the sustainability case. Better explanations can deepen engagement and make an unfamiliar racing format easier to follow. Formula E must ensure that predictions remain clearly labelled, timely, and accurate enough to increase understanding rather than distract from the race.
The expanded multi-year agreement follows a partnership first formalised in January 2025, suggesting the organisations saw enough value to broaden deployment. Yet net-zero claims require transparent measurement. Formula E should quantify which travel, equipment movements, energy use, and event changes were avoided, including the footprint of the digital infrastructure itself.
The most compelling part of the strategy is that carbon efficiency is treated as an operating constraint, not a separate reporting exercise. When digital twins reduce travel, AI optimises energy recovery, and better coordination prevents waste, sustainability and performance can improve through the same decision. That is a more durable model than using AI to describe emissions after they have already occurred.
