What operational technology includes
Examples include programmable controllers, industrial supervisory systems, building automation and devices that measure or act on a process. Unlike an administrative application, their operation can directly influence movement, temperature, pressure or ventilation.
NIST emphasises that OT protection must consider physical safety, reliability and performance requirements. The distinction between IT and OT helps organise responsibilities, although the two are often connected. An inventory must show these relationships: knowing only a machine’s name does not explain which services or access it needs to operate.
How a risk to people can arise
Scenarios include loss of supervision, incorrect signals, unauthorised parameter changes or unavailable auxiliary systems. Consequences depend on the design, independent safeguards and the task under way. An IT failure does not automatically mean an accident, but the ways it could contribute to one must be examined.
Ask specific questions: what happens if communication is lost? Could unexpected movement start? What information does the operator need to stop? How is a safe state maintained? The analysis must connect the digital incident with observable physical conditions.
Who needs to work together
Operations understands process conditions; maintenance identifies equipment and failure modes; technology and cybersecurity teams review communications and access; prevention contributes the assessment of harm and task organisation. Suppliers may also hold essential information on limits and recovery.
Joint guidance from ACSC and other authorities prioritises people’s safety and recommends understanding critical systems and their dependencies. Define responsibilities and a channel for reviewing decisions: connecting a machine, enabling remote support or replacing a component can change risk even if its appearance remains unchanged.
Controls and changes to plan
Measures must address the actual architecture and criticality. Common considerations include network separation, authorised access, configuration protection, recoverable backups and control of third-party connections. No generic checklist can establish that a particular facility is protected.
In OT, updating or scanning equipment without preparation can interfere with its operation. Coordinate intervention windows, tests, acceptance criteria and rollback arrangements with competent personnel and manufacturers where appropriate. The urgency of a vulnerability must be managed alongside process safety, rather than through improvised production changes.
Incident response and safe recovery
The plan must explain how to report a suspected incident, who decides whether operations continue and how workers at the facility will be protected. Technical containment decisions must consider their physical effects. Disconnecting a component without understanding its dependencies can create additional problems.
Recovering files or restoring communications does not by itself demonstrate that a process is safe. Verify configurations, required functions and start-up conditions against previously agreed criteria. Preserve evidence and review causes. Management must link to the site’s emergency measures and preventive risk assessment.
Hypothetical example: remote support for a facility
A plant proposes remote access so a supplier can review its ventilation system. Before enabling it, the plant identifies which components could be changed, who will authorise each session and what consequences parameter changes could have during work.
The team defines limited permissions, traceability and a procedure for checking changes. It also agrees how to maintain safe conditions if the connection fails. The project is not approved simply because it has a password: the relationship between access, operation and protection of people is validated, and recovery is rehearsed without exposing staff.
