Audi is working on smart manufacturing with AI

The traditional assembly line is running up against its limits. Its rigid nature and fixed lead times no longer match, for example, the growing variety of versions of a model and all the possibilities of individualisation. To be more flexible and achieve greater efficiency, Audi is therefore working on making the 'Modular Assembly' ready for production. This is happening with a pilot at the Ingolstadt plant. This revolves around the assembly of a door panel for the interior.

In the pilot, work does not happen according to a fixed pattern. Tuning is possible with regard to individual requirements and wishes. AGVs (automated guided vehicles) bring the door panels to the station where assembly takes place. This can be linked to specific, additional production steps. This modular method enables one employee to also assemble awnings, for example. With the traditional production line, this still requires 2 or 3 employees within the fixed lead time. This kind of 'Smart Production' can improve productivity by 20 per cent. Another important advantage here is that this method is often also suitable for employees who cannot work (any longer) on a traditional assembly line due to physical limitations.

To get new designs and parts into production as quickly as possible, Audi uses powerwalls. These are huge screens on which cars are displayed in full size and extremely detailed 3D. Even lighting and shadows are highly realistic. These generated views allow important design decisions to be made at an early stage. Thanks in part to special glasses, Audi experts can experience a 3D model of a car from the customer's perspective. This too helps to shorten the time between design and production. 3D simulations are also used to further improve quality. For example, they quickly reveal where any tolerances can be expected.

Digital assembly planning also plays an important role in the smart manufacturing of the future. Here, a digital scanning process generates three-dimensional 'cloud points'. These can be used to fully visualise, control and adjust machines and infrastructure in the factory. This uses software based on artificial intelligence and machine learning. This process eliminates the need to build prototypes in the planning process. This saves resources and enables flexible cooperation between different sites.

Spot plays a key role in the scanning process. This robot dog independently determines its route through the factory and along the production lines. He is capable of fully scanning the Neckarsulm factory in 48 hours. This factory is where the Audi A6 is built. A human takes three weeks to do this. This innovative process makes it possible to build new Audi models virtually first. Any pain points are quickly revealed in the simulations, before actual production begins.

To save energy, Audi is mainly using data and data analytics. The in-house developed Energy Analytics tool, for example, clearly shows energy consumption at times when there is no production. This includes the consumption of machines, for example, but also lighting, ventilation and heating. The clear overview enables targeted interventions. Thus, Audi saved around 37,000 MWh of energy by 2021 thanks to Energy Analytics.

Artificial Intelligence (AI) and machine learning are playing a crucial role in the digital transformation of Audi's manufacturing process. For example, an AI algorithm together with an artificial neural network helps detect imperfections in components. The basis for this is deep learning. This is a form of machine learning where huge amounts of data create patterns. No deviation from this is allowed. In another pilot, Audi is deploying artificial intelligence to measure the quality of welding points. This makes it possible to automatically monitor in real time the 5,300 welding points required when assembling the body of an Audi A6. Previously, quality control was done randomly.

For the benefit of factory automation, Audi is deploying the Edge Cloud 4 Production concept. In this, a few centralised local servers replace large numbers of (expensive) industrial PCs. This has significant advantages. Previously, the introduction of new software and control programmes and functions, for example, always required new hardware. With the new concept, new software suffices. Audi has successfully tested this centralised server solution in practice at the Böllinger Höfe plant, where small-scale series production takes place (Audi R8 and e-tron GT quattro). The aim is to roll it out further at all Audi Group plants.

In the field of digitalisation, Audi is working closely with the Technical University of Munich and the Fraunhofer Institute for Industrial Engineering and Organisation. This takes place on the Heilbronn campus. With the Automotive Initiative 2025 (AI25), Audi aims to establish the network for digital factory transformation and innovation. As a driver of ideas, AI25 plays an important role in the digital transition. The Audi factory in Neckarsulm counts as a pilot plant in this respect. As such, this production facility sets the tone for the digital transformation of the entire Volkswagen Group. To identify new and innovative ideas and integrate them perfectly into production, Audi is also using the Production Lab in Gaimersheim.

Source and photo: Audi