The Volkswagen Group is one of the world's leading automotive manufacturers. The group includes twelve brands from seven European countries, including Audi, SEAT, ŠKODA, Bentley, Bugatti, Lamborghini, Porsche, Ducati, Scania and MAN. During vehicle production, Volkswagen Group brands must affix 25 different labels that contain country-specific safety, usage and specification data - with more than 2,000 variants. Damaged, incorrect or missing labels can lead to legal problems. In addition, errors can occur during inspection, as employees are not always able to read the labels to verify accuracy.
To automate this process, an intelligent sign inspection (iSI) solution on Amazon Web Services (AWS) was developed for Volkswagen's Porsche brand. This solution is powered by the Digital Production Platform (DPP), a cloud-based industrial platform developed on Amazon Web Service to enable the development of scalable use cases across the Volkswagen plant and transform manufacturing and logistics processes. The Digital Production Platform (DPP) will also serve as the foundation for the Industrial Cloud - an open platform, community and marketplace for digital solutions available to the automaker.
Volkswagen follows strict quality control processes to ensure that every vehicle label is correct and undamaged. However, plants produce vehicles that are sold worldwide, so employees must deal with labels in multiple languages. To achieve the goal of zero label defects, Porsche, a Volkswagen brand, had an automated solution developed for tag inspection.
"iSI helps workers reduce the complexity of inspection tasks and helps Porsche further improve vehicle quality," says the product manager for artificial intelligence-based computer vision at Porsche.
The iSI solution is based on Volkswagen's Digital Production Platform, which is driving the use of Internet-of-Things (IoT) solutions to innovate supply chains. The DPP is powered by AWS IoT Core, which enables developers to connect IoT devices to the cloud without having to provision or manage servers.
"We benefit from the DPP Connect team providing standardized IoT solutions for our factory assets,"
explains the product owner for artificial intelligence-based computer vision at Porsche. Volkswagen's Industrial Computer Vision (ICV) Camera Service, for example, uses AWS IoT Core to connect cameras on the factory network to the cloud.
Porsche uses iSI to automatically check labels in the factory with IoT-enabled devices and to perform manual checks with a mobile app. On the assembly line, automatic cameras photograph the labels and send them to the cloud for analysis. If a problem is detected, the system notifies a worker on the assembly line to take the vehicle to a quality control station for manual inspection.
During the manual inspection, workers photograph individual labels using the iSI app. Using Amazon Textract - a fully managed machine learning service that automatically extracts printed text, handwriting and other data from scanned documents - the iSI app compares the text on the labels with data stored in Volkswagen's ICV Inference Service to detect anomalies. The results are translated and displayed to the employee, who can take appropriate action to correct detected errors.
By using AWS services to build iSI, Volkswagen has moved a step closer to its zero defects goal. In the future, the group plans to expand the solution to all of its global plants. "Our iSI use case is just the first of many more to come," explains the product owner for artificial intelligence-based computer vision at Porsche. "With our modular approach and standardized services, we can transfer the iSI solution to other brands in the Volkswagen Group and use the ICV functionalities to implement further use cases for the entire production logistics."