Automation ≠ Intelligence
For decades, manufacturing has pursued automation — replacing human labor with machines. But the real competition ahead is not about who owns more equipment — it's about who possesses a production system capable of continuous optimization. AI is driving manufacturing from "execution automation" to "autonomous optimization" — enabling equipment to perceive changes, understand its own status, and co
Why are many factories becoming increasingly automated, yet still not truly intelligent?
This is the biggest shift in manufacturing over the next decade.
Recently, capital has been intensively focusing on Physical AI.
Hangzhou Lianhui Technology completed a multi-hundred-million-yuan financing round and launched a multimodal model for the physical world.
Many people are focused on: model parameters, technical metrics, and financing amounts.
But what truly deserves attention is: this signals that manufacturing is entering a new phase.
Over the past few decades, factories have pursued: automation.
Letting machines replace repetitive human labour.
Robots replacing manual handling.
Equipment automatically completing processing.
Production lines running automatically.
But automation has a ceiling: machines are merely executing rules pre-written by people.
Who sets the programs? Who adjusts the parameters? Who judges anomalies?
Still people.
So many factories look very advanced. Lots of equipment. Lots of systems. But when changes occur, they still require significant human intervention.
The next step for true smart manufacturing is not about making machines execute faster.
It is about making machines begin to optimise autonomously.
What does this mean?
Take a machining tool, for example.
In the past: engineers set processing parameters, and the equipment runs according to the program.
In the future: the equipment can sense in real time: material changes, tool conditions, environmental changes, and processing results.
Then it adjusts its own parameters – improving quality stability, efficiency, and energy consumption.
This is true intelligence.
The biggest change that Physical AI brings is enabling AI to enter this closed loop for the first time.
AI of the past: responsible for analysing information.
AI of the future: responsible for influencing reality.
It needs to: see the world, understand states, make judgments, and take action.
01From Automation to Autonomous Optimization
This is also why the future of industrial competition is not just about who has more equipment.
It is about whose production system has stronger autonomous optimisation capabilities.
Because the biggest cost in manufacturing is not just labour.
An even greater cost is: uncertainty.
Changes in equipment status, changes in product demand, changes in materials, process fluctuations.
Truly advanced factories are not those without change.
They are those that can adapt to change quickly.
02The Opportunity for Industrial Software: From Management to Evolution
This is also the greatest opportunity for industrial software in the future.
The software of the past helped enterprises manage production.
The software of the future will help production systems evolve continuously.
Of course, this path is not easy.
Because the industrial world is not like the internet.
There is no unified data, no standardised environment, no simple feedback loops.
It requires connecting: equipment data, process knowledge, quality results, and production experience – all together.
Forming an industrial intelligence system that can continuously learn.
So the manufacturing competition over the next decade may see a new watershed:
Ordinary factories: have automated equipment.Advanced factories: have production systems capable of autonomous optimisation.
A truly intelligent factory is not about machines replacing people.
It is about enabling the entire production system to possess the ability to continuously evolve.
This may be where Physical AI truly transforms manufacturing.