Case Study: FEV Europe GmbH finds critical driving situations faster with Microsoft Azure Stream Analytics

Day-in, day-out: finding the needle in a 40-terabyte haystack

FEV Europe GmbH, an automotive engineering services provider in Germany, needed a way to analyze up to 40 terabytes of sensor data per vehicle per day from test fleets developing SAE Level 3 autonomous driving systems. The goal was to identify the most critical driving situations quickly and reliably so engineers could validate and train automated driving software more efficiently. Microsoft Azure Stream Analytics, along with Azure IoT Hub, Azure Data Lake Storage, and Azure Key Vault, was used as part of the cloud-based approach.

Microsoft Azure enabled FEV to build a secure, turnkey data pipeline that ingests vehicle signals, detects critical events in near real time, and automatically marks relevant time stamps for later extraction and labeling. The solution stores only critical-situation data in hot storage while moving other raw data to cold archive storage, reducing cost and manual effort. According to FEV, the platform helped streamline development significantly by making it possible to focus human review on only a small fraction of the overall data and by speeding up validation for fleet-wide autonomous driving tests.

FEV Europe GmbH

Markus Kremer

Project Engineer