How Porsche Cup Brasil Uses AI for Race Management

Porsche Cup Brasil has deployed an AI-powered crash analysis system to accelerate race operations. When a car crashes, engineers photograph the damage from multiple angles and upload the images to a web app running on Azure Kubernetes Service. A Python backend routes the photos through an AI multi-agent in Microsoft Foundry, which identifies damaged components from a catalog of about 2,000 parts. Engineers review and validate the AI's output, and corrections are fed back to improve the system. The tool, developed with Microsoft partner Kumulus, has cut the time needed to assess and start repairs by roughly half. In addition, real-time telemetry from onboard sensors is streamed into Microsoft Fabric and visualized in Power BI dashboards, enabling engineers to detect anomalies and call cars into the pits or stop them to prevent failures. Future plans include a garage scheduler agent to automate parts ordering and a data agent to combine crash analysis with telemetry. Organizers emphasize that AI remains a decision-support tool, with engineers retaining final control.