The biggest industrial AI opportunity may be between the tools, not inside them.

One cylinder. Several versions of the same knowledge.

Take a simple example: a pneumatic cylinder.

It first appears in the mechanical design. Its characteristics are already known: part reference, stroke, sensors, location in the machine and function.

Some of this information is then re-entered or reconstructed in electrical engineering to define its connections and inputs and outputs (I/O). It is used again in automation engineering to define its behaviour, then in the human-machine interface (HMI), alarms, documentation, simulation and sometimes, later, maintenance.

Speeding up each discipline with better tools addresses only part of the problem.

A shared representation of the machine.

This is a question we are currently exploring at FACTREN.

Specialist environments such as SolidWorks, EPLAN, TIA Portal and DELMIA would retain their roles. The idea is to connect their contributions through an engineering layer that spans disciplines.

Such a layer could maintain a shared representation of the machine, progressively enriched as each discipline makes its decisions.

• Mechanical engineering supplies the facts.
• Electrical engineering adds its decisions.
• Automation engineering adds the behaviour.
• Simulation can contribute the sequence.

That knowledge could then be reused to generate or inform several deliverables, instead of remaining confined to individual software environments.

A broader role for industrial AI.

This also changes the potential role of AI.

Alongside generating PLC code or schematics more quickly, AI could help interpret, structure and retain the engineering reasoning that connects the different disciplines.

The value would extend beyond automating an individual task.

It would lie in using the same information or engineering decision several times, after defining it once.

The opportunity between disciplines.

For software developers and teams building industrial AI, we see a particularly interesting area to explore: the interfaces between engineering disciplines.

For industrial companies and automation integrators, the question could shift from:

“How can we automate more work within each engineering department?”

towards:

“How can we stop our engineering knowledge from being continually recreated as it moves from one discipline to another?”

This is precisely the kind of question FACTREN wants to explore with industrial technology developers and the people who use their tools in the field.