⦁ Audi brings artificial intelligence (AI) to Audi design for the first time
⦁ Software brings brand new inspiration to the creative process
⦁ By keeping AI software development entirely in-house, Audi demonstrates expertise in a critical new area
Ingolstadt, December 12, 2022 – The use of artificial intelligence (AI) in all areas: That is the goal that Audi has set itself on the way to becoming a data-driven company. With FelGAN, the company is now using AI-supported software to open up new sources of inspiration for designers.
Creative people are always looking for inspiration. The same applies to the designers who create new wheels in the Audi Design Studio in Ingolstadt. But where do you find untapped sources of inspiration?
The principle of “thinking outside the box” is well known and catchy. But it’s more difficult to implement because people are more likely to draw on what’s familiar to them in the creative process. This is exactly what the AI-supported FelGAN software, which was developed by Audi IT and Audi Design, is intended to achieve. Now the project allows creative people to use an almost unlimited amount of ideas. Interacting with the software allows designers to discover motifs from completely new perspectives and gives them the opportunity to further develop and edit their creations. In concrete terms, FelGAN works by either quickly offering a large number of photorealistic designs or by combining existing designs in a targeted manner.
In this way, the system acts as a hub of spontaneous ideas for the Audi wheel design team to exchange new versions and variants. The tool allows designers to easily experiment with shape, color, surface texture and other parameters in real time.
How artificial intelligence learned to design wheels
The name “FelGAN” is a mixture of the German word for “rim” (rim) and “GAN”, the latter being an acronym for Generative Adversarial Networks. GAN is a special form of a self-learning computer program in which two algorithms compete against each other in so-called learning and become better and better in the competition.
It works like this: One of two algorithms, a “generator”, creates artificial images of a certain subject – in the case of FelGAN – a car rim. The discriminator – a competitor, so to speak – sees a selection of images consisting of photos of real wheels and images from the generator. Now the discriminator decides whether each image is a creation of the generator or a real photo. This process is repeated over and over again until the workout is complete.
Both algorithms are designed to learn from their mistakes and constantly improve. After enough passes, the creations of the generator are so deceptively real that even the human eye can hardly or not at all distinguish them from real photographs.
Intuitive application user interface based on Streamlit technology offers short development cycles and quick feedback between designers and IT professionals. The computing-intensive components of the AI application run in the cloud so that developers are not dependent on powerful local hardware when operating a software solution.
Cooperation between humans and AI
Another benefit of FelGAN is that the software assigns a mathematical value to each design generated by the AI. These values, called “DNA” by the developers, can be used at any time to reproduce the design. But that’s not all: Audi designers can also upload their own drawings and photos to the program and add them to the virtual experimentation area. This is based on special algorithms that determine the appropriate DNA values for images entered by designers.
Designers often only use individual elements from FelGAN creations and process them into a harmonious overall design. In addition to mastering manual tools, a creative eye and professional experience play a decisive role here. Finally, the Audi experts turned the virtual design into reality by using a high-tech milling machine to produce a prototype wheel made of plastic or aluminum.
Innovations from the world of Audi
FelGAN was developed and implemented entirely in collaboration between Audi IT and the design department. The company is thus demonstrating its expertise in the field of software and in the future field of artificial intelligence.
Thomas Knispel, Head of Machine Learning & Data Science at Audi: “In the modern age, data bring immense added value for companies and their employees. Audi has committed to the goal of becoming a data-driven company. To this end, we are going to use AI in many departments.
So our data team is always on the lookout for new technologies.”
In the future, the technology behind FelGAN could be expanded into a comprehensive AI design platform that could also serve as an inspiration for designers from other Audi departments. In addition, an AI scoring system is currently being developed that evaluates each FelGAN rim in terms of its carbon footprint. In summary, FelGAN is a step that Audi has taken on the way to becoming a data-driven digital company.