News, 16 April 2025
Porsche Engineering is making ADAS (Advanced Driver Assistance System) validation more efficient by using its ComBox App and Peregrine.ai’s object detection to allow any test vehicle to contribute to traffic sign recognition validation.
Challenge: Validating ADAS driving functions requires extensive measurement data from test drives to cover various scenarios, including construction sites and unusual traffic signs . Traditional measuring equipment is expensive, making it costly to scale across large vehicle fleets, and the validation process is labor-intensive, often involving manual comparison of traffic sign recognition by a developer.
Solution: To reduce manual effort, Porsche Engineering is using smartphones with traffic sign recognition software from Peregrine.ai. Porsche Engineering provides the interface to a data logger, connection unit, and camera hardware, while Peregrine.ai integrates its AI-powered object detection. The smartphone compares the vehicle’s recognition of traffic signs with Peregrine.ai’s object detection results.
Process: When discrepancies occur, the smartphone captures the traffic sign and instructs the data logger to record relevant vehicle data. The ComBox App enables Porsche Engineering to receive data from in-vehicle systems, allowing for cloud-based comparison and ADAS function optimization. This method reduces manual labor, optimizes traffic sign recognition, and can be cost-effectively deployed across more test vehicles.
Future Applications: In the future, smartphones in test vehicles can detect and record errors, not just in dedicated ADAS optimization vehicles, to capture more edge cases and improve ADAS quality, especially in regions like Australia with unique traffic rules. This approach, already tested during the development of the new Porsche Macan, may be used in series development and extended to optimize other ADAS features like lane detection, traffic light phase detection, and pothole detection.
