The hype surrounding Artificial Intelligence in engineering is massive. Almost every software vendor now offers smart AI features. When applied correctly, AI-powered solutions deliver maximum leverage right where engineers and designers waste valuable time every day, whether on tedious routine work or background tasks. Examples include automated test case generation, understanding complex documents, or quickly assessing minor design changes.
But behind the scenes, the industry’s reality paints a sobering picture. According to a CIMdata study, while about 80% of software vendors already offer AI features in their portfolios, the actual adoption rate among industrial customers is only 8% to 33%. Nearly nine out of ten industrial companies use AI in less than a quarter of their projects, leaving most initiatives stuck in pilot mode. So why is adoption stalling?
The challenge: the invisible data wall
The biggest hurdle isn’t the algorithm itself, but what comes next: AI integration. For industrial customers, seamless connection to existing systems is a key selection criterion. Yet software and service providers systematically underestimate this barrier. In practice, service providers encounter unexpected issues with legacy systems 2.4 times more often than customers anticipate. When new AI tools are bolted onto existing systems in isolation, they fail when faced with the reality of complex engineering processes. As our Chief Product Officer Frank Patz-Brockmann puts it: “The real pain point isn’t buying the software but getting it activated.”
Integration over API patchwork
CONTACT Software is tackling this problem head-on with Fourier AI. We need to view AI system integration as a fundamental architectural decision, not an afterthought or a quick DIY project. Instead of connecting an isolated external AI via standard interfaces, Fourier AI is deeply embedded as a fully integrated intelligence layer within our proven CONTACT Elements platform.
In customer discussions, we sometimes encounter an urge to throw established data structures overboard and replace them with a “data lake” or even a “data swamp” in the hope that the AI will somehow piece the information together on its own. However, this is entirely the wrong approach, because AI thrives on structured data.
That’s why with Fourier AI, we rely on a clear, bottom-up layered architecture:
A Single Source of Truth (PLM) as the foundation: At the very base is still the PLM system with its highly structured data. This foundation is essential and serves as the reliable starting point for any intelligence.
Context building: Simply storing data somewhere isn’t enough. A precise context must be constructed for the AI. For example, if a user submits a prompt like “Give me all data on this Engineering Change,” this context layer needs to know exactly which data belongs to it, who created it, and how everything is connected.
The model layer: Sitting above that is the model layer. In addition to leading external models, we offer our own specialized models, such as for 3D similarity search. The system automatically selects the right model for specific platform tasks while giving companies the freedom to choose which models they want to access for their custom use cases.
AI orchestration & governance: To evaluate and continuously improve answers, generated execution traces are captured and logged. The system filters data so the model only returns information the specific user is authorized to see via integrated permission controls.
User interaction (the PLM chatbot): At the top layer is the user, who interacts directly by typing a query into the PLM chatbot, for example.
Conclusion: Foundation first, value follows
This deep architectural approach initially requires more diligence when structuring data and preparing systems. But that’s precisely where the key advantage lies: because the foundation is cleanly built from day one, the AI layer automatically adapts whenever a customer expands their data model in CONTACT Elements – without writing a single line of additional code. This is how we bridge the gap between mere tech gimmicks and genuine, productive value in day-to-day engineering.
By trade, I am a UX designer. My heart beats for intuitive interfaces, consistent user experiences, and the question: How does a user get from A to B as easily as possible? For years, the answer in business software was always fairly straight forward: clear menus, learnable patterns, and efficient paths through the system. But is this question still relevant given the impact of AI on UX?
Artificial intelligence doesn’t just change what software can do. It changes how we work with software. This fundamentally reshapes our definition of what a ‘good user experience’ actually means.
The end of the classic user interface?
Let’s look at an everyday engineering example: You receive an email. Your supplier informs you that component #4711 will no longer be available starting in March. Previously, you would click through the bill of materials (BOM), manually search for where it is used, research alternatives in another system, create a change request, and notify colleagues via email. All of this takes hours, scattered across multiple tools.
Soon, you will only have to say: “Component #4711 is discontinued – what is affected and what are the alternatives?” The system independently checks all assemblies, suggests qualified replacement parts, and prepares the change request for approval.
Take it one step further into the future, and you won’t even have to ask. The system detects the impact itself, proactively suggests the change, and simply waits for your approval.
H2: New requirements for AI UX
This example illustrates the fundamental shift that AI capabilities and agents are currently ushering into engineering environments. It’s a transformation that upends much of what we’ve learned about UX.
The AI UX of tomorrow will no longer be measured by how intuitive an interface is to use. Instead, it will be measured by:
…how well the system understands the user’s intent
…how relevant the proactive suggestions are
…how intelligently the automation works in the background
Simply put: The best UX will probably be the one that requires the least amount of interface interaction to get the job done.
Five principles redefine UX
What does this mean in practice? I see five shifts emerging—especially in the engineering environment.
1. Conversation instead of clicks
Voice and text are becoming the primary interface. Not because clicking is bad, but because natural language is closer to our actual intent than any menu structure. A well-trained system understands what is meant—even if the phrasing isn’t perfect.
This requires a radical rethink: It is no longer about users learning how to navigate the interface. The system must learn to understand humans.
2. Proactive instead of reactive
Today’s software waits for input. Future systems act on their own: they identify risks before they arise, suggest actions before the user looks for them, and point out inconsistencies before they turn into problems.
A quality agent that monitors data streams and automatically initiates corrective action – this is not science fiction. It is a concrete architectural decision that can be made today.
3. Adaptive workspace – just in time
The interface configures itself based on role, task, and data context. A design engineer needs a different interface than a project manager or a quality assurance manager. And what you need at 9:00 AM might be completely different from what you need at 3:00 PM.
Intelligent systems adapt to the human – not the other way around.
4. Knowledge graph instead of keyword search
Classic search finds documents containing a keyword. Semantic search based on a knowledge graph understands relationships: Where is this component used? Which variants are affected? What changed in the last version?
This difference sounds technical, but it has immediate UX consequences: users no longer need to know exactly what to search for. The system delivers the right context.
5. Human-in-the-loop
As impressive as AI agents are, decision-making authority remains with humans – and it always should.
Good AI UX means finding the right moment for the system to pause and wait for human approval—not too early (which would be annoying), and not too late (which would be risky).
Human-in-the-Loop is not just a safety net. It is a design decision that builds trust.
A fleet of digital colleagues
One analogy helps me when I think about the future of working with AI: Imagine that instead of getting a new tool, you are getting new colleagues who take care of engineering and design, who specialize in quality and compliance, and keep an eye on the supply chain.
These digital agent-colleagues work silently in the background, orchestrated and coordinated with one another. They are not omnipresent, but they are always there when you need them.
This changes not only the work itself, but also our relationship with software. A tool that must be operated becomes a partner that thinks ahead.
AI assistants integrated into CONTACT Elements support engineers with their daily tasks
What this means in practice
These changes won’t happen overnight, and they are not a given. Anyone who wants to benefit from intelligent UX needs the right foundation: structured data and clear governance. And a system that views AI not as an add-on, but as an integral component.
The quality of AI outputs is directly dependent on the quality of inputs. An agent accessing unstructured or incomplete data will deliver unreliable results and poor UX, no matter how well-designed the interface is.
That is why my key takeaway from my work over the last few months is: Investments in data quality and platform architecture are investments in User Experience.
Putting people at the center
AI changes the interface, but not the objective. Technology should empower people to do their jobs well. With less friction, more context, and stronger foundations for decisions.
The fact that this will happen in the future via conversation instead of clicks, proactive suggestions instead of empty text fields, and semantic connections instead of keyword searches is not a threat to good UX design. It is an invitation to rethink what “intuitive” actually means.
Deploy AI the right way
Artificial intelligence only delivers real impact when AI features are systematically and valuably integrated into your processes. Discover how this works in our AI Insights. This newsletter provides you with tangible real-world examples, best practices, and valuable tips on all aspects of engineering and manufacturing intelligence.
Artificial Intelligence (AI) is already an integral part of modern engineering. Especially when you‘re dealing with the development of complex machinery and its major challenges. For example, long product lifecycles, safety, compliance, and service traceability, decades of legacy product knowledge and multi-discipline engineering (mechanical, electrical, software, controls). In these areas, AI can help to reduce complexity, and thus, increase development speed and time-to-market.
Tools like ChatGPT and GitHub Copilot support many areas across a wide range of tasks. While these applications are easy to use, successful AI engineering is far more complex. This is due to intricate processes as well as a frequent lack of data quality, quantity, and infrastructure.
PLM software not only provides a platform for managing product data. It’s an important interface for implementing AI-powered technologies across various engineering phases.
How to implement industrial AI in engineering?
Model training and data provisioning remain key hurdles for successful industrial AI.
Ready-to-use AI applications like ChatGPT and DALL-E are available for general purposes, such as text processing or image generation. However, there is a lack of instantly usable applications or models for complex industrial issues. Publicly available datasets rarely meet specific industrial requirements. As a result, you need to fine-tune AI models and build company-specific datasets and custom models.
In addition, deploying AI models in industrial environments presents further complexities. This includes
the integration with existing systems,
access security and authorization when managing the underlying data as well as
ensuring reliability and performance.
PLM systems provide the infrastructure to collect, structure, and process data. They facilitate AI model integration into existing processes and offer a central platform for managing AI solutions throughout the entire product lifecycle. This includes ensuring data quality, version control, and managing the generated outputs.
By connecting AI tools and PLM systems, your company ensures a consistently structured use of data. Thus, you built the foundation for successful AI applications.
A second major challenge lies in the complexity of AI implementation itself. This ranges from the selection of suitable models and algorithms to the adaptation of these models to the specific requirements of your company. To make AI engineering work, IT, engineering, and data science must collaborate closely. PLM systems facilitate this by providing a unified platform that bridges these disciplines and drives teamwork.
PLM systems as enablers for AI
PLM systems with deep integration capabilities create the necessary conditions for incorporating AI into existing development processes. As central data hubs, they enable access to data from various sources in the product lifecycle.
By providing a consolidated data foundation, PLM systems clear the path for generative AI (GenAI) technologies. These advanced solutions independently generate fresh, creative content, from text to images, offering companies unprecedented opportunities, particularly within design and product development.
By facilitating the integration of AI-powered design tools, PLM systems ensure efficient management and versioning of AI-generated designs. For instance, AI-powered tools can automatically generate design variants of CAD models. They can also optimize these designs based on simulation results.
Connecting AI solutions to existing data structures is another critical aspect. By integrating AI into PLM systems, AI-powered design tools can be linked directly to your product data. This accelerates the entire development process. This improves component reuse, enhances collaboration between teams, and continuously optimizes product development.
Furthermore, PLM systems capture AI-generated insights and optimizations in a central repository, making them available for future use. As a result, you can significantly increase your capacity to innovate.
AI-based applications in engineering
The interaction between GenAI and PLM systems enables the integration of AI-driven design processes into existing workflows. This unlocks a wide range of applications:
Automated design variants
AI creates new design variations based on existing CAD data and optimizes them for parameters such as cost, material usage, or production options. This gives your engineers more time for creative design tasks.
Simulation-driven optimization
The synergy between simulation tools and GenAI algorithms enables a continuous feedback loop, allowing design concepts to be iteratively refined and optimized. This helps you to identify the best design choices based on comprehensive data analysis. It reduces the number of physical prototypes and saves both time and costs.
Optimized product development
AI dynamically adjusts designs to new requirements during the ongoing development process. In connection with the PLM system, your company can react faster to changes and adapt product development without delays. This offers significant benefits, especially for mechanical and plant engineering.
Time-series forecasting
GenAI analyzes time-series data to predict future developments. By leveraging historical and operational data, PLM systems combined with AI help you to identify trends at an early stage and make informed decisions.
Data management and deployment
Integrating AI into industrial engineering is not only a technical challenge but also an organizational one.
Successful AI deployment requires suitable data management. PLM systems provide a platform to manage data in a structured manner – from collection and storage to provision for AI models. Data protection, security, and availability must be guaranteed.
In addition to data processing, AI deployment – the availability in the production environment – is also crucial. To integrate AI, your IT and engineering teams must work together. PLM systems make this process secure and efficient, while ensuring employees can easily use AI in their daily work.
Therefore, PLM systems must support the development and integration of AI models as well as ensure their operation and maintenance in day-to-day business.
PLM systems also serve as a platform to monitor, update, and improve AI models. By using operational data, these models can easily adapt to changes, ensuring consistently high product quality.
Trends and perspectives
PLM systems will increasingly serve as central platforms for managing and integrating AI solutions. AI-powered automation will become more prevalent, as the ability to learn from data and recognize complex relationships is constantly improving.
Another trend is the growing integration of AI in collaboration across different departments and companies. PLM systems will develop into platforms that support both internal processes and seamless cooperation across company boundaries. This leads to more efficient supply chains and accelerated innovation. Additionally, explainable and transparent AI models are becoming increasingly important to strengthen trust in AI and increase acceptance in safety-critical areas of engineering.
Success factors for AI in engineering
Data availability and quality
High-quality data and its availability form the basis for the success of AI applications. PLM systems ensure that the required data is available in the appropriate quality through consistent data collection and management. This allows models to train on a solid database and deliver precise and reliable results.
Seamless process integration
To exploit the full added value of AI technologies, you must integrate them into existing workflows. PLM systems integrate AI-powered applications into existing processes and simplify their use. As a result, AI solutions can be implemented into existing systems without major adjustments and put into operation more quickly to deliver relevant results immediately.
Training and change management
Companies must prepare their employees specifically for the use of the new tools by providing technical training and conveying an understanding of their use in the respective work context. Well-planned change management promotes acceptance and actively involves all employees in the change process.
Management support
Support from management is crucial for the successful introduction of AI in engineering. Communicating clear goals and strategies for the AI transformation is just as important as providing the necessary resources. Meanwhile, management must foster a culture of change.
Conclusion
The integration of AI is revolutionizing engineering. PLM systems play a key role as the central hub for data. They create the necessary infrastructure to integrate AI applications into existing processes and use them efficiently. This enables you to optimize your product development processes and exploit the full potential of AI.
The path to success lies in the sensible use of data, the integration of AI models into the existing IT landscape, and the involvement of employees. Only those who tackle these challenges in a targeted manner will be able to shape the AI transformation in engineering sustainably.
A strategic approach is key. Companies should focus equally on technology and people. This is how they sustainably unlock the full potential of artificial intelligence in engineering.
Webcast: Accelerate complex machinery development
Ready for real results? If your team’s AI projects are still struggling with messy data or adding unwanted complexity to your product development, it’s time for a reality check. Watch our Industry Talk to discover what it takes to make AI deliver reliable value for your engineering team.
What you will learn:
The Pitfalls: Why most AI initiatives fall short in complex product data environments.
The Foundation: The non-negotiable foundation needed to deliver truly reliable AI insights.
The Integration: Practical strategies to seamlessly embed AI within your existing engineering workflow.
