Guide to Operational Technology (OT) Security — NIST SP 800-82 Rev. 3

NIST SP 800-82 Rev. 3 defines operational technology broadly as programmable systems and devices that interact with the physical environment. Its durable value for the bank is the way it separates functional control-system families while showing the same basic loop underneath them: sensing, controller decision, actuation, operator interaction, and maintenance.

What carries weight #

The architecture material in sections 2.3.1–2.3.4 is the useful core. It distinguishes geographically dispersed SCADA supervision, plant-wide distributed control, and smaller PLC-centered systems without pretending that the categories are mutually exclusive. It also emphasizes constraints that shape real designs: safety, timing, geographic distribution, availability, environmental conditions, and the need to preserve local control when higher layers or communications fail.

The guide is especially useful for correcting two weak mental models. First, SCADA is not a synonym for every industrial HMI: it is a supervisory system built around remote assets and communications. Second, a PLC is a controller component that can appear inside a SCADA or DCS architecture or serve as the center of a small independent system.

Where not to lean on it #

• The document is a cybersecurity guide, not a control-engineering design standard. Its topology diagrams establish vocabulary and attack-surface context, not latency budgets, redundancy requirements, or a bill of materials.
• The examples are intentionally generic. A real design still needs process hazards, control-loop dynamics, failure modes, availability objectives, communications constraints, and applicable sector standards.
• The familiar level or zone picture describes responsibilities and trust boundaries; it does not prove that any message path is deterministic, safe, or authorized to command the process.

Visual audit #

Pages 27–39 of the PDF were rendered and inspected at readable scale. Figures 1–8 carry the architecture argument: the control loop, SCADA control-center/field-site layouts, point-to-point and multi-drop communication topologies, rail SCADA, DCS, and PLC-centered examples. They are clear as source evidence but too security-specific and visually dense to reuse at article width. The topic entry therefore uses an independently drawn role-and-authority diagram verified against these figures, while the prose retains the important distinctions among local control, supervision, operations, and enterprise consumers.