CIE vs. Traditional OT Security: What Changes When Engineers Own the Problem
Traditional OT security tries to keep the adversary out. Cyber-Informed Engineering assumes they get in and asks what the plant can still survive.
By Timothy Porritt · Porritt Inc. · November 21, 2026
Ask a security team and a process engineer how to protect a compressor station and you will get two different answers. The security team will talk about segmentation, access control, and monitoring. The engineer will talk about surge protection, relief valves, and trip logic. Traditional OT security formalizes the first answer. Cyber-Informed Engineering formalizes the second and connects it to the first.
Different Starting Assumptions
Traditional OT security, as captured in frameworks such as IEC 62443 and the NIST guidance for industrial control systems, organizes the plant into zones and conduits, assigns security levels, and applies controls to keep unauthorized actors and code out. Its implicit goal is to prevent compromise and detect it quickly when prevention fails.
CIE begins from the assumption that a determined adversary will eventually achieve some level of access. Its goal is to make sure that access cannot be converted into an unacceptable physical consequence.
Side by Side
| Question | Traditional OT security | Cyber-Informed Engineering |
|---|---|---|
| Starting point | Networks, assets, vulnerabilities | Physical consequences and critical functions |
| Core assumption | Compromise can be prevented or caught | Compromise will happen at some point |
| Primary owner | Security and IT/OT teams | Design, process, and controls engineers |
| Typical controls | Firewalls, authentication, patching, monitoring | Mechanical limits, hardwired interlocks, simplification, manual modes |
| When applied | Often after the system is designed | From concept and design basis onward |
| Measure of success | No intrusion, fast detection | No unacceptable consequence, even with intrusion |
Why the Network-First Approach Runs Out
Network defenses age. A firewall rule set that was tight at commissioning accumulates exceptions. Patching is slow in plants that run for years between outages. And the most serious industrial attacks on record did not rely on exotic flaws in perimeter devices; they used engineering tools and control functions for their intended purpose, in the wrong hands.
The 2017 attack on a petrochemical facility’s safety instrumented system is the clearest case. The target was the last automated layer of protection. Where a plant’s only defense against a hazardous scenario is a programmable system reachable from an engineering workstation, a security failure becomes a safety failure.
Where Engineering Runs Out
CIE has limits too. Not every consequence can be engineered away at reasonable cost, and existing plants cannot be redesigned from scratch. Engineered barriers do nothing to protect data, detect an intruder, or support incident response. A plant with excellent mechanical protection and no monitoring will survive an attack without ever learning it happened.
How They Fit Together
The two approaches are layers of one defense. CIE decides which consequences get a non-digital barrier and reduces how much the remaining digital systems can do. Traditional security then protects what is left, with its effort focused where the consequence analysis says it matters most. In practice the consequence analysis often shrinks the security problem by removing functions and connections nobody needed.
Who Should Be in the Room
The organizational change is the hard part. CIE asks process, mechanical, electrical, and controls engineers to treat malicious action as a design input, and asks security staff to understand the process well enough to know which signals matter. For the principles behind this, see The INL CIE Framework Explained; for the basics, What Is Cyber-Informed Engineering?.
Related reading
- What Is Cyber-Informed Engineering? A Plain-Language Guide for Engineers
- The INL CIE Framework Explained: The 12 Principles in Working Terms
- A Cyber-Informed Engineering Implementation Checklist for Plant and Utility Teams
- Cyber-Informed Engineering for Utilities: Designing the Grid to Fail Safely
Timothy Porritt is founder of Porritt Inc., building AI-powered tools for process safety, engineering compliance, and industrial operations. Based in Salt Lake City, Utah. Questions about applying Cyber-Informed Engineering to a specific facility: timothy@porrittinc.com.


