What is OT/ICS Security?
Everything else in this course has, implicitly, been about IT — information technology, the systems that process and move data. OT (Operational Technology) is a different world entirely: the systems that directly control physical processes — a power grid's circuit breakers, a water treatment plant's chemical dosing pumps, a factory's robotic arms, a pipeline's pressure valves. When OT security fails, the consequence isn't a data breach. It's a blackout, a contaminated water supply, or a physical explosion.
:::
IT vs. OT: not the same problem wearing a different hat
| IT | OT/ICS | |
|---|---|---|
| Primary goal | Confidentiality, then integrity, then availability (classic CIA order) | Availability and safety first — the process must keep running safely |
| Typical lifespan | 3–5 years, replaced routinely | 15–25+ years — a control system installed in 2005 may still be running today |
| Patching | Regular, often automatic | Rare, scheduled around planned outages — sometimes years apart |
| Downtime tolerance | Usually acceptable in planned windows | Often near-zero — stopping a power plant or a hospital's systems has real-world consequences |
| Consequence of compromise | Data theft, financial loss, reputational damage | Physical damage, environmental harm, injury, or loss of life |
| Network exposure | Often internet-facing by design | Historically isolated ("air-gapped"), increasingly connected anyway |
Notice the ordering flip in the very first row. IT security training drills "Confidentiality, Integrity, Availability" (CIA) as the priority order. OT security effectively runs it backwards — a control system that's perfectly confidential but stops the process from running safely has failed at its actual job.
Why OT security lagged behind IT security
Several factors compound to make OT genuinely harder to secure than IT, not just differently prioritized:
- Equipment predates modern security thinking entirely. Many industrial protocols were designed decades ago, when the assumption was "only trusted engineers on a physically isolated network will ever talk to this," with no authentication or encryption built in at all.
- You cannot simply reboot a power plant to patch it. A patch that would take an IT server offline for five minutes might, on OT equipment, require a planned outage scheduled months in advance, coordinated with regulators and customers.
- "It's air-gapped" was a comforting assumption that stopped being true years ago. Modern OT environments increasingly connect to IT networks for monitoring, remote support, and business reporting — the gap most organizations believe still exists usually has more bridges across it than anyone realizes.
- The people who understand the physical process (control engineers) and the people who understand security (IT/security teams) have historically been different teams, with different vocabularies, reporting up different chains.
:::warn "Our OT network is air-gapped, so we don't need to worry about this" is one of the most common and most dangerous assumptions in this field. Real-world incidents have repeatedly traced back to a "temporary" remote-access connection, a contractor's laptop plugged in for maintenance, or a USB drive — any of which quietly bridges a gap everyone assumed was absolute.
The vocabulary you'll need
- ICS (Industrial Control Systems) — the umbrella term for systems that monitor and control physical processes
- SCADA (Supervisory Control and Data Acquisition) — systems that monitor and control processes spread across a large geographic area (a power grid, a pipeline network)
- DCS (Distributed Control System) — systems that control a process concentrated in one facility (a single power plant, a single factory)
- PLC (Programmable Logic Controller) — the actual "brain" device that reads sensors and controls physical outputs (valves, motors, relays) at the equipment level
- HMI (Human-Machine Interface) — the screen an operator actually looks at to monitor and control the process
These aren't interchangeable jargon — SCADA and DCS describe different architectural scopes, and PLCs are the hands-on hardware that ties software decisions to physical reality. Later lessons in this path go much deeper on each.
A concrete example of the stakes
Consider a water treatment plant that uses a control system to regulate how much chlorine gets added to the water supply. In IT terms, this is "just" a control loop reading a sensor value and adjusting an output. In OT terms, if an attacker (or a bug, or a misconfiguration) manipulates that control loop to add far more chlorine than intended, the consequence isn't a corrupted file — it's contaminated drinking water reaching real people. This exact category of attack — remote manipulation of a water treatment control system — has happened in real incidents, which is precisely why this field treats "worst case" so much more seriously than most IT security training does.
Where this path is going
- SCADA and industrial control system fundamentals (deeper architecture, components, and how a control loop actually works)
- Common industrial protocols (Modbus, DNP3, and why their design assumptions cause security problems)
- The Purdue Model — the standard reference architecture for segmenting IT from OT
- Real-world ICS attacks and what they teach us (Stuxnet, the Ukraine grid attacks, and others)
- Practical OT security controls that respect the "don't break the physical process" constraint
- A capstone exercise applying the Purdue Model to a realistic industrial network
🧪 Hands-on lab — reframe an IT incident as an OT incident
Pick a well-known IT security incident you already understand (a ransomware attack, a data breach — anything from earlier in this course).
Rewrite the scenario as if it happened to an OT environment instead — what physical process would be affected, and what would the worst-case consequence actually be?
Compare your OT version's worst case to the original IT version's worst case. Which one is more urgent to prevent, and why might a security budget still (wrongly) favor IT anyway?
Write one sentence explaining why "we'll just patch it during the next scheduled downtime" is a much bigger constraint in OT than it typically is in IT.
Explain, in your own words, why OT security effectively reverses the classic CIA triad's priority order.
Why is "our OT network is air-gapped" often a false sense of security in practice?
Name the difference between SCADA and DCS in one sentence each.
Key terms
- OT (Operational Technology) — systems that directly monitor and control physical processes and equipment
- ICS (Industrial Control Systems) — the umbrella term for OT systems and their supporting infrastructure
- Air gap — a network with no physical or logical connection to other networks, historically assumed (often incorrectly) to protect OT environments
- PLC (Programmable Logic Controller) — the hardware device that directly reads sensors and controls physical equipment
Recap
OT/ICS security is a genuinely different discipline from IT security, not IT security applied to different equipment — the priorities, timelines, and consequences are fundamentally different, and treating OT like slow-moving IT is exactly how real incidents happen. The rest of this path builds out the architecture, protocols, real-world attack history, and practical controls that respect what makes OT different in the first place.
References
- CISA — Industrial Control Systems — the US government's primary ICS security resource hub
- NIST SP 800-82 — Guide to Industrial Control Systems Security — the foundational technical reference for this entire path