Why Control Engineering Skills Matter More as AI Expands in Automation

Artificial intelligence is making it easier to generate code, analyze documentation and automate parts of the control-system workflow. That does not reduce the importance of control engineers. Instead, it shifts more value toward process knowledge, engineering judgment, system-level thinking and the ability to understand how physical equipment behaves outside normal operating conditions.

Automation World contributor G Brooks-Zak argues that the hardest part of controls engineering has never been typing ladder logic or structured text. The harder task is deciding what the machine should do, why it should do it and how it should respond when reality does not match the original design assumptions.

AI Can Generate Logic, but It Does Not Define the Process

Modern AI tools can produce PLC code, suggest software structures and assist with documentation. Those capabilities can reduce the time required to create routine logic.

However, a control program still represents a physical process. Motors accelerate and decelerate. Valves have response times. Sensors fail. Conveyors jam. Materials vary. Operators intervene. Maintenance teams introduce temporary workarounds that sometimes remain for years.

A control system that ignores those realities may compile successfully and still perform poorly in production.

Process Knowledge Becomes More Valuable When Coding Gets Faster

As code generation becomes easier, the engineering bottleneck moves upstream. Someone still has to define operating sequences, interlocks, permissives, alarm behavior, recovery logic and safe responses to abnormal conditions.

That requires understanding both the automation platform and the equipment being controlled. It also requires knowing which variables are truly important and which apparent software problems are actually mechanical, electrical or process problems.

For experienced controls engineers, this combination of disciplines is often what makes troubleshooting effective. A recurring fault may originate in timing, sensor placement, mechanical wear, electrical noise or process variation rather than in the PLC program itself.

Failure Modes Matter More Than Normal Operation

Normal operation is usually the easiest part of a machine to describe. The difficult engineering work appears around edge cases.

• What happens if a sensor changes state late?
• What happens if a motor starts but feedback never arrives?
• How should the line recover after an emergency stop?
• Which devices must remain energized during a controlled shutdown?
• What should an operator see when two faults occur together?
• How should the system behave when communication to another controller is lost?

These conditions are difficult to solve through generic code generation because they depend on the specific machine, process, risk and operating environment.

System-Level Thinking Remains a Core Controls Skill

A controls engineer has to understand relationships across the complete system. PLC logic interacts with mechanical motion, drives, safety systems, networks, HMIs, upstream and downstream equipment, maintenance procedures and operator behavior.

A change that improves one part of the system may create a problem elsewhere. For example, shortening a sequence may increase throughput but reduce settling time. Changing an alarm threshold may eliminate nuisance alarms while hiding an early warning. Increasing an acceleration rate may save cycle time while increasing mechanical stress.

AI tools can help analyze these decisions, but they still require accurate context and engineering review.

Good Communication Is Part of Technical Competence

Controls engineering is also collaborative. Operators often know how a machine behaves after years of production. Maintenance technicians know which faults repeat and which components fail first. Process engineers understand material variation and quality constraints.

Extracting that knowledge requires good questions and clear communication. It also requires translating practical experience into sequencing, diagnostics, alarm messages and recovery procedures.

These skills become more important when AI is introduced because an AI system also depends on clear context. Engineers must describe the task precisely, review the generated result and determine whether it makes sense for the actual machine.

Industrial AI Raises the Importance of Data Quality

AI-assisted control engineering also depends on the quality of the information available to the model. Incomplete asset data, outdated drawings and disconnected production records limit what an AI system can infer.

This connects directly to the broader manufacturing-data problem discussed in our analysis of manufacturing data connectivity. If design, production, quality and maintenance data remain isolated, AI will also inherit those gaps.

The same challenge appears at the platform level. Our Siemens industrial AIoT article examines how industrial platforms are attempting to contextualize equipment and process data before applying analytics and AI workflows.

AI Should Be Treated as an Engineering Tool, Not an Authority

The strongest use of AI in controls engineering is likely to be as an engineering assistant. It can accelerate repetitive programming, summarize documentation, identify patterns and generate first-pass logic.

The engineer remains responsible for validating the result against physical constraints, safety requirements and operating objectives.

This is especially important in industrial environments because incorrect logic can affect equipment, product quality, uptime and personnel safety. Faster code generation does not reduce the cost of a wrong engineering assumption.

Skills That Become More Important in AI-Assisted Automation

• Process knowledge: understanding how equipment and materials behave in real operation.
• Failure-mode analysis: anticipating abnormal conditions and safe recovery paths.
• System-level thinking: understanding interactions between controls, mechanics, networks and operations.
• Troubleshooting: separating software faults from electrical, mechanical and process causes.
• Communication: extracting useful knowledge from operators, maintenance and process teams.
• Validation: reviewing AI-generated logic against engineering requirements and physical constraints.
• Data literacy: understanding whether the information used by AI is complete, current and correctly contextualized.

The Engineer’s Role Is Changing, Not Disappearing

AI may change who writes the first version of a function block or sequence. It does not remove the need for someone who understands the machine.

As programming becomes faster, engineering judgment becomes a larger share of the project’s value. Controls engineers who combine automation expertise with process understanding, communication and system thinking are therefore likely to become more important as AI capabilities expand.

Source and Original Author

Original author: G Brooks-Zak
Original publication: Automation World, September 21, 2026
Original article: The Control Engineering Skills That Matter More as AI Capabilities Expand

G Brooks-Zak is co-founder of Outlier Automation and a member of the Control System Integrators Association. This AITUOS Controls article is independently rewritten and expanded for technical context.