Industrial Intelligence

70% localization — the rules of the game have changed.

As the target for domestic substitution of industrial software rises from 35% to 70%, what truly deserves attention may not be the word "substitution" itself — but the fact that AI is reshaping the R&D paradigms and competitive rules of industrial software. In the past, we were playing catch-up with the previous generation of products. In the AI era, does China's industrial software have a real op

Domestic industrial software substitution rate is set to rise from 35% to 70%.

What does this mean?


For decades, Chinese industrial software has been playing catch-up.

CAD, CAE, PLM, EDA – it is not that there are no domestic players, but rather that they have often been competing within rules already defined by others.

Why is this so difficult?

Because industrial software is not ordinary software.

Behind it are mathematics, algorithms, physical mechanisms, modelling and simulation, plus decades of engineering validation.

So the real barrier has never been about whether code is written fast enough.

It is about:
whether you have truly turned industrial knowledge into software capability.

This is why many domestic industrial software products appear to have increasingly comprehensive features, yet still struggle to fully replace foreign counterparts in high-end industrial scenarios.


01A Critical Variable Is Emerging

But now, a critical variable is emerging.

AI is beginning to change how industrial software is developed.

In the past, developing industrial software involved engineers searching for information, writing code, building models, tuning parameters, and performing verification.

What might the future look like?

Engineers articulate requirements, AI understands those requirements, invokes industrial knowledge and mathematical models, generates code and solutions, calls simulation tools for verification, and finally the engineer confirms.


Note:This is not about adding an AI chat interface to industrial software.It is about changing how industrial software itself is developed and used.

This is why the recent AI-Empowered Industrial Software Conference in Suzhou is particularly noteworthy.

The conference proposed that "AI-native architecture will lead the next generation of industrial software systems," and simultaneously launched the AI4I initiative, beginning to close the loop between requirements, development, verification, and application.


02Where Does the Opportunity Lie for Next-Generation Industrial Software?

So the question arises:

"What exactly will the next generation of industrial software look like?"

And what is the real opportunity that AI brings to domestic industrial software?

If the rules of the game remain unchanged, Chinese enterprises can only continue chasing.

Foreign software has accumulated decades of advantage, and you must spend decades to catch up.

But if AI fundamentally transforms software development methods, knowledge organisation, and engineering interaction models, then the leadership accumulated over the past decades may not be carried forward intact.

So what is truly worth expecting from domestic industrial software this time is not simply creating a "domestic version of a certain foreign software."

Rather, it is:
whether we can, with the help of AI, redefine industrial software itself.


03True Integration Capability Is the Key

Of course, this path will never be easy.

AI can help you write code, but the core mathematical models, physical mechanisms, engineering rules, and verification systems of industrial software still require long-term accumulation.

So the truly competitive companies in the future may be neither pure AI companies nor traditional software companies.

Rather, they will be those that can truly integrate:
industrial knowledge, mathematical models, AI capabilities, and real engineering scenarios.

This is also why there is now growing emphasis on pilot verification, scenario openness, and industrial ecosystems.

Because industrial software ultimately is not about delivering a demo – it is about:
whether it can actually run in a real industrial environment.


Therefore, the most noteworthy aspect of the 35% to 70% figure is not simply "how much the substitution rate has increased."

Rather, what it implies is:


Chinese industrial software is moving from "catching up with the previous generation" to "competing for the right to define the next generation of industrial software."And AI may be the greatest opportunity in this lane change.