One of the most visible developments at the event was Eplan Copilot. It brings together engineering knowledge, platform functions and specific tasks through natural-language interaction. For example, an engineer looking for the projects in which a particular component has been used can enter its part number. The AI searches the projects stored in the cloud and returns the results.

“We want to make it easy for users to quickly get their bearings in the Eplan software and provide them with information and functions exactly when they need them,” Fleming said.

Eplan draws a clear distinction between offering assistance and taking over tasks autonomously. Initially, Copilot will use verified Eplan expertise. Knowledge from partner companies, customers and standards will be added in the future.

“In engineering, the results must be reproducible and reliable. For this, the architecture has to be right,” Fleming said. Copilot will not rely on information scraped indiscriminately from the internet. “Even a machine that is designed 99 per cent correctly still won’t work when it comes down to it.”

The vision remains ambitious. With a growing knowledge base and additional skills, what starts as a copilot is intended to develop gradually into an autonomous pilot.

Eplan Platform 2027, available since the beginning of September, marks the next stage of development. The company says around 500 customer requests were incorporated into the development process. Fleming describes the goal as “unconditional user-friendliness”.

New functions in Eplan Preplanning, Eplan Electric P8 and Eplan Pro Panel are designed to simplify and further automate planning steps, including preliminary planning and 3D project generation. Fleming sees the release as more than a conventional software update.

“The core benefit that Eplan provides companies is the ability to scale—that is to say, making the completion of more projects and more variants manageable using the available resources,” he said.

This includes organising reuse systematically. Rule-based project generation can already deliver significant productivity gains without necessarily requiring AI.

“Scaling doesn’t start only with artificial intelligence,” Fleming said. “When we standardise, reuse and automatically assemble functions, we gain speed and quality, while also addressing the shortage of skilled workers.”

Much of this process already works through rules. “In Eplan, standardised, reusable schematic functions are assembled into project-specific solutions. AI doesn’t play much of a role in this”—yet, he said, significant economies of scale are still achieved.

Cloud-based data management is also becoming more strategically important. Project, master and device data can no longer be treated as separate entities. With Eplan Platform 2027, synchronising master and device data through the Eplan Cloud is set to become a more integral part of the workflow.

“For us, end-to-end means that data doesn’t stop at the perimeter of a single application,” Fleming said. The aim is therefore to create more than a user-friendly interface: the platform is intended to become a consistent, comprehensive data foundation for subsequent process steps.

At Eplan Next26, manufacturer Eaton showed how far this data foundation can extend into downstream processes. At a switchgear-system plant in China, the Eplan Platform serves as the backbone for electrical engineering. Humanoid robots use cameras to inspect wiring and check whether labelling matches the engineering documentation.

Fleming sees this as more than a striking AI use case. “The robot scans the cabinet, compares the labelling with the schematics and recognises incorrectly labelled or incorrectly wired connections. This demonstrates the advantages that clean engineering data can deliver, all the way through to quality assurance.”

Eplan Smart Sourcing, which also premiered at Next26, demonstrates how these processes can work. The solution connects electrical engineering with procurement. During planning, engineers can check whether required components are available from manufacturers and distributors connected to the system, and when they can be delivered. Critical parts can be replaced earlier, and validated bills of materials can be passed to the procurement department.

“Smart Sourcing is a great example of how we’re closing the gap in the end-to-end process. It connects a technical decision made in engineering directly to actual availability in the supply chain,” Fleming said.

Such connections depend on participating systems speaking the same language, which is why open standards are central to Eplan’s strategy. “AI isn’t worth much without clean, standardised data,” Fleming said.

The Asset Administration Shell is designed to make data transferable in a structured way, regardless of manufacturer. ECLASS Advanced provides an important semantic foundation. The Industrial Digital Twin Association (IDTA) uses the AAS to define the structure, interfaces and semantics for a standardised digital twin.

For Eplan, this is about more than standardisation. The Data Portal currently contains more than four million static and configurable device specifications from more than 550 manufacturers. When such data is available in a standardised format, it can be used, reused and processed automatically—not only in schematics but across additional process steps.

“Data can only become a true driver for productivity when it is unambiguous and can be reused without the need for translation or reformatting,” Fleming said. In his view, the cloud, AI and AAS are not separate technology trends but building blocks of the same end-to-end architecture.

Fleming acknowledges that the AAS has been discussed for years. In Germany, he says, there is extensive conceptualisation and scientific analysis, but pragmatic implementation in marketable solutions does not always happen quickly enough. He sees an opportunity to build the necessary critical mass.

“Just committing to a standard won’t get us anywhere. The key question is whether or not we can create solutions that are easy for people to use,” he said. Large industrial companies need to build real-world applications, rather than simply pay lip service to the AAS.

Product Change Notification (PCN) shows how this approach can work in practice. Fleming describes the challenge from a component manufacturer’s perspective: “A component is delivered to a control cabinet manufacturer, the control cabinet is installed into a machine and this is then delivered to the operator. In the end, as the component manufacturer, how do I reach the user of my component when I have no idea where it’s been installed?”

The AAS and Eplan Projects can connect these data points. Eplan and Phoenix Contact have developed an AAS-based use case for this purpose. If a manufacturer changes or discontinues a component, or issues a safety-relevant update, the information can be transmitted in a structured way and processed in the Eplan Data Portal. The engineer receives a notification and can identify the projects or existing systems using the affected item.

“These are exactly the applications where we see whether a standard creates real added value or not. The information shouldn’t just sit somewhere, unused. It must automatically get to the place where it triggers a decision,” Fleming said.

At the SPS trade show in November, Eplan and Phoenix Contact will present the new PCN solution. “We’re demonstrating that in Germany, we can take a really good idea and establish a pragmatic solution on the market,” Fleming said.

This development is also changing Eplan’s role. The company is moving from supplying an engineering platform towards becoming a key player in a digital ecosystem connecting component manufacturers, machine and plant-system builders, control cabinet manufacturers, distributors and operators. The priority is not only the functions the software offers, but whether data remains usable across companies and processes.

This brings the focus back to Eplan Next. At the Connected Industry panel at Eplan Next26 in May, participants committed to developing concrete solutions using the AAS. SPS will mark the next milestone, when Eplan and Phoenix Contact present their first solution.

The next Eplan Next has been announced for June 9 and 10, 2027, in Cologne. Fleming promises further product innovations and a larger presence, as well as visible progress on data-driven topics.

“Eplan Next27 will be a proof point for us. Have we made significant strides with AAS and end-to-end processes—or will we be standing there, one year on, saying we really ought to?” he said.

Fleming sees two lines of development converging. AI, automation and improved usability are intended to help users become the “best-possible engineers”. Meanwhile, standardised data and open interfaces aim to connect the process chain across organisational boundaries.

Autonomous engineering would not be the starting point, but the result of a well-structured data and process architecture. “Only when the data is consistent across the board can we truly use artificial intelligence consistently,” Fleming said.