AI use gains pace at Audi

AS ARTIFICIAL INTELLIGENCE (AI) moves from experimentation to execution, the race is on to see who can capitalise first (and the most) on the burgeoning technology.

AI’s rapid adoption is no more evident than in the automotive space exemplified by premium German brand Audi which is fast tracking its use in car production and other allied areas.

Essentially, the four-ringed brand is using AI to increasingly control production systems via a large cloud instead of stationary factory computers at scale and has formed partnerships with institutions such as IPAI to complement Audi’s “innovation strength”.

“Artificial intelligence is a quantum leap for efficiency in our production,” said Audi production and logistics board member Gerd Walker.

“With our AI and digitalisation roadmap, we are transforming our plants into smart factories where AI acts as a partner, providing our employees with tailored support.”

The first AI-controlled robots are taking over tasks that are ergonomically strenuous, and chatbots are providing additional relief.

“We are bringing together Audi’s decades of production expertise, our own innovative strength, and the expertise of strong partners such as the Innovation Park Artificial Intelligence (IPAI) in Heilbronn,” the company said in a statement.

Reports from sources at Audi-central, the German cities of Ingolstadt and Neckarsulm, say the automaker is consistently rolling out AI in production and logistics while introducing new AI use cases and technologies for large-scale series production.

Audi says it is putting the Edge Cloud 4 Production (EC4P) into operation across its production environment and in doing so, is “setting the next benchmark” in fully networked factory automation while at the same time laying the foundation for the widespread use of AI in production.

Cleverly, EC4P makes it possible to combine conventional automation technology with flexibility and computing power from the cloud.

It allows Audi to simplify processes, reduce the amount of hardware required on site, and introduce new functions more quickly making processes more stable, reducing maintenance costs, and increasing IT security.

As a practical example of AI implementation, Audi cites vehicle assembly at its German plants where worker guidance is increasingly being controlled centrally from the cloud.

“Employees on the production line receive information such as vehicle specifications or regional versions from a central source in real time. The move to the cloud has already eliminated the need for more than 1000 industrial PCs,” commented Audi.

A further example of AI implementation is in Audi’s A5 and A6 body shop in Neckarsulm where EC4P is now being used in a highly automated production environment for large-scale series production for the first time. 

Virtual programmable logic controllers (vPLC) are replacing local hardware controllers on the production lines.

Industrial devices, including around 100 robots, work together via the EC4P with millisecond precision.

The EC4P meets the highest requirements for a smooth production process, enabling several hundred vehicle bodies to be manufactured every day in three shifts – a benchmark that is unique in the industry to date.

AI is also being deployed after concerns for employee safety, and quality assurance, addressed by systems such as the incoming Weld Splatter Detection (WSD) system that will also run on the EC4P in the future, enabling even greater flexibility and scalability.

The WSD detects weld splatter on the underbody of a car body and marks it with light then a robot arm takes over grinding them down – a physically demanding job.

The first AI-supported weld spatter detection system will soon go into series production at six plants in Ingolstadt.

Audi is also developing ProcessGuardAIn, its own AI solution for monitoring manufacturing processes.

It was made possible by a team of Audi data experts who have built up the cross-plant “P-Data Engine” platform over the past few years.

Audi says the platform combines various system and plant data from production at a uniform quality level.

“Thanks to this database, data scientists at Audi can quickly and efficiently develop and scale AI applications – such as ProcessGuardAIn,” added Audi.

“The AI solution bundles decades of expert knowledge as well as plant and process data in a standardised, scalable modular system that can be used across the Volkswagen Group.”

Looking further down the track Audi is developing the Next2OEM project working with 10 partners at its Ingolstadt headquarters to demonstrate how the production and assembly of a wiring loom can be completely digitised and automated – from the supplier to installation in the factory.

To date, less than 10 per cent of wiring loom production and assembly is automated across the industry.

And the use of AI at Audi even extends to the paint shop with an AI-based dryer operation.

On its path to data-driven (AI) production, Audi is utilising its own know-how and the expertise of strong partners from industry and science.

Company resources include around 60 experts in the Audi Production Lab (P-Lab) and the P-Data Factory who are driving new technologies forward—from the initial idea to large-scale production.

Partners include Broadcom, Cisco, and Siemens, with Audi itself implementing the interaction between virtualisation platform, network, and automation technology as part of EC4P.

Additionally, Audi is an active partner in the IPAI in Heilbronn, the European hotspot for applied AI, since 2023.