AI is change management - at the speed of light

His lecture titled "Speed of light in engineering through AI" was an open account of technological decisions, organisational prerequisites, and cultural factors that, in his view, determine the success or failure of AI projects.

What actually is AI - and what isn't?

At the beginning of his lecture, Zein cleared up a common misconception: Not every software that calls itself "AI" meets these criteria from a technical perspective: "Essentially, AI is fundamentally machine learning. And you can say: If it’s not machine learning, it’s not AI."

Machine learning means recognising patterns in data to make reliable predictions. Zein explained this with a simple example: Only when sufficient data is available can a pattern be recognised - and that is precisely the core of AI. Building on this are deep learning approaches with neural networks and so-called foundation models like large language models, which are now known as "generative AI."

"They generate new content from existing content - and that's what we currently see with ChatGPT and similar systems."

Why automation is a necessity

With an open view of the current challenges facing the industry, Simon Schmidt, senior manager business unit automated intralogistic, manufacturing and assembly systems from Fraunhofer IPA, opened the Automation NEXT Conference 2026. He addressed the question of how automation has developed, what forms it takes today, and what role AI-based systems will play in the future.

Philip Bittermann

Ai in electrical CAD: From circuit diagram to finished control cabinet

WSCAD develops electrical CAD software for planning electrical systems - from machines and test stands to building automation. This is exactly where the company's use of AI comes in. Zein described a typical work process in electrical design: after creating the circuit diagram, the complex planning of the physical control cabinet follows - a work step that can take several hours to a whole working day, depending on experience.

With the AI-supported solution like the one from WSCAD, this process can be largely automated. “You simply mark your pages and say: Generate cabinet. What happens in the background: the system analyses the circuit diagram, recognises all components, selects the appropriate mounting plate, pulls DIN rails, places the components - and after two minutes the result is there.”

The designer remains in control of the process at all times. Iterative adjustments, manual placements of individual components or specific requirements for certain control cabinets are still possible. Wiring, routing and checking the cable duct utilisation are also automatically taken into account.

The effect: "Usually, you need six hours for such a thing - now you're done in two minutes."

From vision to market-ready solution

What Zein explicitly emphasised: The described functions are not theoretical future visions. WSCAD began development in 2023, launched the first version in 2024, and is now presenting the second generation of the software. Customer reports from practice show clear effects: "My project duration has been reduced by 50 percent," quoted Zein from a user. Translated, this means: With the same team, almost double the performance can be achieved.

This development has not gone unnoticed outside the customer base. WSCAD has been awarded, among other things, an award from the specialist magazine Schaltschrankbau for AI technology and as an innovative medium-sized company.

AI development as a top priority - and room for experiments

The development history of the AI functions is deliberately described by Zein as a "top priority." An initial team received the clear mandate to develop a proof of concept for an AI co-pilot within two months. In parallel, a second development team worked without an official mandate on an alternative approach - inspired by models like the "20 percent time" from Silicon Valley.

Both approaches led to usable results. However, a key insight was realised early on: "The added value arises when AI supports business goals."

In the case of WSCAD, this means allowing electrical designers to work faster, more efficiently, and with less stress - especially the 80% who do not spend several hours a day in the system and do not know complex menu structures by heart.

A striking example is the automated bill of materials creation: where competitor software requires numerous clicks, a simple command suffices here. "Generate BOM - done. An apprentice can do that. You don't need to employ a designer for it."

Lessons learned: What AI projects really need

Zein dedicated a central part of his presentation to the "lessons learned" from his own development - and the typical mistakes companies should avoid:

• Data quality: "You need clean data. Start today, not tomorrow."
• Coalition of the willing: Small, motivated teams with expertise and user perspective are crucial.
• Speed and transparency: Small teams, clear goals, fast iterations.
• Error culture: "We discarded our original model eight times." An approach that is hardly conceivable in many organisations.
• Focus on customer benefit: Technical sophistication without clear added value interests no one in the end.

Zein clearly classified AI as a form of change management - but "at the speed of light". Technologies that seemed impossible a year ago are now easily implementable. Companies must learn to deal with this dynamic.

Corporate culture as a competitive advantage

Why AI has worked at WSCAD, Zein explains less with technology and more with corporate culture. Performance is measured not only by results but also by behaviour. The central principles include focusing on the essentials, the commitment to try new things, and the explicit permission to fail.

Collaboration is particularly important: “There is no single person who has an overview of everything today.” Electrical design, fluid technology, process engineering, building automation, and AI must be considered together - not in isolated silos.

Respect plays a central role: Other opinions are expressly welcome, even towards management. “You can also argue with me - but respectfully.”

How IIoT solutions, adaptive worker guidance, and MES are revolutionising production

Christian Berger from Wilo demonstrated at the Automation NEXT Conference how modern IIoT solutions, adaptive worker guidance, and MES systems create efficiency, flexibility, and transparency in production. Learn how Wilo uses these technologies to tackle challenges such as the skills shortage and increase productivity - and how gamification helps motivate employees.

Bernhard Richter

Is AI a job killer? A clear classification

Zein answered the question of whether AI destroys jobs unequivocally, referring to figures from the World Economic Forum: While around 92 million jobs will disappear by 2030, about 170 million new ones will be created - a clear net gain.

More decisive, however, is the qualitative change in work. For designers, AI primarily means relief from routine tasks: searching for parts, inserting macros, creating bills of materials, checking standards, manually following up changes. “The grunt work - nobody wants to do that anyway.”

Instead, the role shifts towards managing and controlling AI-supported processes. Designers would become more productive, have more freedom, and could refocus on what defines their profession.