Enterprise Resource Planning Blogs by SAP
Get insights and updates about cloud ERP and RISE with SAP, SAP S/4HANA and SAP S/4HANA Cloud, and more enterprise management capabilities with SAP blog posts.
Showing results for 
Search instead for 
Did you mean: 


Hardly anyone is not talking about generative AI these days. And many are wondering how this new technology is going to change the professional landscape. In this blog post I am going to speculate what we can expect to change in the near future in relation to Digital Thread in Product Lifecycle Management.

To set the stage, let’s define our terms.

Generative AI refers to a type of machine learning that involves training artificial intelligence models to create new data and output that did not exist in the training data. Unlike classification and prediction algorithms, generative AI models can generate output based on their understanding of the patterns and structures in the training data.

Digital Thread is a concept used in Product Lifecycle Management (PLM) that refers to the digital representation of a product from its conception to the end of its lifecycle. SAP PLM Digital Thread model involves the following stages:

  1. Plan: defining the product requirements, including customer and regulatory requirements, design specifications.

  2. Design: creating and developing the product design, including conceptual design, detailed design, and engineering specifications.

  3. Build: producing the product, including the creation of prototypes, production planning, and actual manufacturing.

  4. Operate: delivering the product to customers and providing ongoing support and maintenance, repair, and replacement.

In general, we should expect the AI to start being relevant at all stages of the process, but to a different degree. To develop a sense to which degree, you can imagine the following scale of activity:

  1. Human Needs: activity unique to human nature which could be described to an AI but would never be internally understood.

  2. Human Creativity: production of physical or literal artifacts from the perspective of humanity itself.

  3. General Creativity: production of artifacts generally appealing to humanity and not based on personal insights into the human nature.

  4. General Logic: activity that is formulaic, based on clear rules and agreements.

  5. Automation Logic: highly formulaic, repetitive activities.

The efforts of software developers in the last 50 years were focused pretty much on the last two points. With more sophisticated software addressing General Logic through its design and workflows. However now it becomes clear that generative AI is capable to dramatically improve both the Automation and General Logic, and somewhat set foot in the General Creativity space.

Having been trained on symbolic input of the contents of the Internet, generative AI contains a snapshot of humanity’s creative efforts. A reflection of our understanding of the world and what we do in it. As such, generative AI cannot be relied upon to derive what is true about the reality, or to speak from the human perspective. The former, of course, could be addressed with time, if we talk about AI agents operating in the outside world, instead of virtual space.

With this in mind, let’s look how different stages of Digital Thread could be affected in the coming years.


Requirements specification on the high level mostly comes from the Human Needs. The idea of a product, its purpose and functions are determined by humans. Coming up with anything radically new yet staying on the level of Human Creativity takes a real person. So, in the farthest end, when everything else is automated, the creative input into the process would remain human-defined.

On the other hand, regulatory specifications are highly formulaic and generally known in advance. A new product that already has existing similar products on the market can have regulatory specifications generated in a straightforward manner.


Conceptual and product designs can benefit from generative AI. Many products are created with a standard design vocabulary defined by manufacturing processes and consumer familiarity. Design variants can be generated by the AI to become a basis for further development. If custom models are trained using proprietary data of specific manufacturers, the whole lines of products can be designed simultaneously based on previous results.

Engineering specifications, again, are mostly based on formal rules and written knowledge. We should expect them to eventually be generated in full volume. This, however, is not something to be resolved soon. Language, and the current image or video-based models do not have a grasp on geometry and material properties, and specialised AI models will have to be developed. Until then generative AI engineering assistance will take place.


How well the AI can be incorporated into the Build stage largely depends on factors other than the AI itself. What we can expect in the short term is improvement in knowledge management, such as production planning and scheduling. A bit further down the line, creation of virtual prototypes would certainly be a good candidate for automation with generative AI. While 3D visualization of engineering data is already supported by SAP PLM, the next steps would include automated quality analysis and optimization of designs for cost and production efficiency.

Real prototypes and manufacturing itself will require further progress in robotics.


Some of the best applications of generative AI should be expected in customer support. If the model is trained on the data throughout the entire production process, it will have the full knowledge of the product and be capable to assist the customers with any common issues. Creating schedules on maintenance and replacement, organisation of deliveries, and eventually marketing and sales can be improved with the AI.


For many years we were promised the AI revolution, and the signs are – it is coming now. The next 2-5 years will be transformative in how we do our work. Though we should also expect a lot of inertia, as no process can be changed overnight, and there is rarely a good reason to break what already works. But we will see much more of generative AI in all aspects of our lives. I hope this post provided you a good summary of what this change means to Digital Thread and I’m especially excited to see how SAP Enterprise Product Development will adopt this technology.

If you found this post useful, please share your thoughts with us in the comments section and click the “Like” button. Follow the tag SAP Enterprise Product Development to receive notifications on future content in this area. You can also post your question on the SAP community using the tag SAP Enterprise Product Development. Join-in to the social media conversations and engage with us using the hashtags – #SAPEnterpriseProductDevelopment #EPD #SAPPLM


SAP notes that posts about potential uses of generative AI and large language models are merely the individual poster's ideas and opinions, and do not represent SAP's official position or future development roadmap. SAP has no legal obligation or other commitment to pursue any course of business, or develop or release any functionality, mentioned in any post or related content on this website.