3 questions for Uwe Wieland

Prof. Dr. Uwe Wieland is Senior Director Software Product Development & AI-driven Solutions at Volkswagen AG

Copyright: © privat
Munich, 07 August 2025

Especially for medium-sized companies with limited resources, it is crucial not to perceive Industrie 4.0 purely as a technological challenge, but rather as a strategic organisational process. The key lies in starting in a structured way with manageable means – precisely where operational pain points and strategic impact intersect.
The first step is to clearly identify the operational issue with the greatest leverage and to make it assessable using concrete metrics. This, in turn, requires companies to engage with their existing data and to invest specifically in its quality.
A practical entry point is often through simple AI applications that build on these existing and quality-assured corporate data.
A critical lever is the ability to distinguish between non-generative AI and generative AI – and to apply them in a targeted manner. While Non-GenAI (also referred to as “analytical AI”) is designed to analyse existing data – for example through classification, prediction, or pattern recognition – GenAI aims to generate new content based on data: texts, images, code, or even initial robotic processes.
In particular, medium-sized companies can begin with analytical AI solutions that use available, high-quality data to better understand relationships within the organisation. This knowledge can then be used to optimise processes (e.g. support processes) and – provided there is high confidence in the data – to predict process parameters (such as maintenance needs). Generative AI can also complement these efforts by, for example, making the contents of internal support and maintenance documents more accessible and providing them via interactive assistance systems. In a more advanced stage of maturity, Non-GenAI and GenAI can be combined into hybrid solution approaches.
The decisive factor is not the extent of the technology used, but rather the ability to truly activate the organisation: aligning people, processes and responsibility structures so that continuous, practice-oriented competence development becomes possible – data-driven, iterative, and scalable.

Artificial Intelligence does not become a competitive advantage by its mere existence – but through a company’s ability to apply it strategically to its own strengths and challenges. What matters is not only whether an AI system works technically, but also whether it is embedded in an organisation capable of using its outcomes effectively.

The following conditions are especially relevant:

Clarity about the problem.
Many AI projects fail not because of the technology, but because the underlying problem was not clearly defined or sufficiently understood. Companies need the capability to consider challenges holistically across people, technology, and organisation. Structuring methods such as narratives or various canvas models have proven effective in systematically qualifying and enhancing use cases.

Organisational integrability.
An isolated AI use case serves no purpose on its own. Real impact only emerges when technology is integrated into existing processes, responsibilities, and data structures. What’s needed is not only digital systems, but also digital readiness: governance, data management, role clarity – in short, an organisation that has learned to work with technology.

Data availability and data quality.
AI thrives on data. Therefore, access to valid, up-to-date and high-quality data is often the real bottleneck. Companies should invest early in data transparency and data maintenance, as even the best algorithms are ineffective if the underlying data is incomplete or biased. Data must be understood and treated as a strategic asset.

Empowerment instead of dependency.
Sustainable AI usage arises when people are empowered to work with technology – not when they rely blindly on “the machine.” Technology communication is thus a key factor: explaining AI, making it comprehensible, and introducing it in a participatory manner builds trust. And only where there is trust can initiative grow – a vital precondition for ongoing learning and improvement in operations.

Building competencies in context.
Rather than relying on general training, AI competencies should be developed directly within the scope of concrete projects. Practice-oriented formats such as masterclasses, university collaborations, or community-based approaches not only promote expertise but also encourage interdisciplinary exchange and sustainable implementation. This way, a learning organisation emerges – one that does not merely consume but actively helps shape.

Conclusion: A competitive advantage does not arise from technology alone. It comes from organisations that are willing to take responsibility, evolve iteratively – and view technology as a tool to strengthen their own excellence.

Industrial applications are not solely about innovation, but above all about reliability, reproducibility, and integration into existing value creation systems. This is precisely where hybrid AI architectures – combining Generative AI (GenAI) and non-generative, analytical approaches (Non-GenAI) – demonstrate their strengths.
While GenAI opens up new solution spaces – for example, through the automated generation of texts, images, code, or even voice commands for robotic processes – Non-GenAI approaches provide the necessary analytical robustness: for quality control, rule monitoring, or data-driven optimisation. Their combination creates systems that are both creative and precise, adaptive and controllable.
For instance, a voice-based process description can be generated via GenAI, which is then simulated, assessed, and optimised using Non-GenAI models (e.g. decision trees, regression models, rule-based systems). In this way, productive, traceable and scalable applications are created that conceal technical complexity behind user-centred interfaces.
In the industrial context in particular, the following holds true: The quality of the architecture is essential – but success ultimately depends on activating the organisation. Only when technological potential is supported by a structurally enabled workforce – with clear roles, responsibilities, and an understanding of data and processes – can sustainable progress be achieved.
Hybrid architectures are therefore not merely a technological necessity – they are a strategic enabler for realistic, effective, and widely accepted Industrie 4.0 solutions.