Agile Vehicle Development With AI and Data

Why is the Big Data Loop so important for the entire automotive industry?

JOACHIM BISCHOFF: All manufacturers are facing the same challenge: Ever shorter development cycles and the desire of customers to always have up-to-date and individualized functions in their vehicles. If you want to be a technological pioneer and successful in the market, you have to focus on customer satisfaction and, in particular, on individualization. This is where the Big Data Loop helps us: It can be used to personalize the customer experience even more. The other brands in the Volkswagen Group also benefit from these developments.

What else are you using the Big Data Loop for?

BISCHOFF: Among other things, to become more agile during development. Our customers expect the latest functions to always be available to them in the vehicle—just as they are used to from their smartphones. Until now, however, the development cycles of the automotive industry and consumer electronics have been very different. The Big Data Loop helps us to become faster. It allows us to transfer data from the vehicles to the cloud, optimize a function using artificial intelligence (AI), and transfer the new status back to the vehicle for validation in just a few minutes. This noticeably reduces the time to market.

What role does this approach play after the vehicles are delivered?

BISCHOFF: Here, the Big Data Loop enables us to continuously optimize vehicle functions even while our customers are using them. They benefit from regular software updates, for example. But one thing is very important to me: We will not turn our customers into beta testers. Before any new software is installed in the vehicle, it will continue to be extensively validated. This is how we ensure high quality and avoid undesirable side effects in other vehicle systems. However, we can reduce the duration and number of tests through the high level of automation in the Big Data Loop.

What are the biggest challenges with the Big Data Loop?

BISCHOFF: The consistency and context of data collection plays a crucial role: With them, in addition to the actual measured values from the vehicle sensors, the boundary conditions such as weather, geography, road conditions, and day and night situations are also important. The evaluation of the data not only makes it possible to improve driver assistance functions, but also to continuously optimize the powertrain or high-voltage battery system, for example.

What role does artificial intelligence play in the Big Data Loop?

BISCHOFF: A central role: It helps us select data and optimize vehicle functions—especially for nondeterministic functions such as cut-in detection. And of course it is the basis of numerous driver assistance systems as well as highly automated driving in general.

What are Porsche Engineering's unique selling points?

BISCHOFF: As a wholly owned subsidiary of an automobile manufacturer, we understand the complete vehicle. This distinguishes us from many Tier 1 suppliers, who are skilled in their dedicated areas. Equally important are our strong software expertise and skills in the latest AI methods. Today, digital topics and e-mobility account for more than half of all activities at Porsche Engineering. This means we can do both: Vehicle and software. We also have tools developed in-house that we use to accelerate development. With our Car Data Box based on an NVIDIA architecture, we can test new features—based on AI methods, for example—in the vehicle. And in all our activities, we make sure that data protection is “Made in Germany.”

New functions have to be adapted to local conditions. Are they trained differently from region to region?

BISCHOFF: In principle, the technologies used should be the same globally. But it is true: We also have to take local peculiarities into account, such as righthand or left-hand traffic. That’s why we use a cloud for the Big Data Loop and train the functions for different markets with different data.

Finally, a personal question: Would you feel comfortable in a vehicle that is continuously optimized even after delivery?

BISCHOFF: Since a final test will take place before every update for the foreseeable future, I'm not worried in terms of quality and safety. On the contrary: Through the Big Data Loop, my car will be able to benefit from the experiences of the entire fleet in the future. This makes my own product progressively safer and safer. That's why my clear answer is: I would feel absolutely comfortable in such a vehicle.

Text first published in the Porsche Engineering Magazine, issue 2/2021.

Text: Christian Buck
Photo: Martin Wagenhan

Copyright: All images, videos, and audio files published in this article are subject to copyright. Reproduction in whole or in part is not permitted without the written consent of Porsche Engineering. Please contact us for further information.

You have questions or want to learn more? Get in touch with us: info@porsche-engineering.de