Techbanq

Offensive security · operational technology

Where a scan that crashes a controller is the incident

On a control network the ordinary security priorities are reversed: keeping the plant running and keeping people safe come before protecting information, and a scanning tool that would be routine on a corporate network can halt a process or trip a safety system. So the method is passive first, active only in agreed windows, and in places it can only be carried out by an engineer standing on the plant floor rather than over a remote connection.

Cybersecurity credentials

  • National Cybersecurity Authority, Saudi Arabia

    NCA registration

    Registered cybersecurity service provider

  • ISO

    ISO/IEC 27001

    ISO/IEC 27001 certified

Certifications & registrations

The conditions

The conditions that make testing a live plant safe to attempt

On a control network, an unsafe test does not produce a bad report; it produces an outage or a safety event. These conditions are the precondition for starting at all, not optional extras.

A named engineer who can halt the test

An operations person with the authority and the means to stop everything within seconds, present for any active work on a live system.

What it prevents: A test that stops before instability becomes an outage.

Agreed windows and excluded systems

Active technique runs only in scheduled windows, and safety instrumented systems are excluded from active testing entirely.

What it prevents: Active work confined to when the plant can tolerate it, and never on the systems that protect people.

Passive-first technique on live networks

Most of the assessment listens and reads rather than sends, because many control devices behave unpredictably when probed with tools built for IT.

What it prevents: Findings gathered without the probe that a fragile controller cannot survive.

A tested rollback for any change

Where a step alters any configuration, the way back is confirmed before the step, not improvised after it.

What it prevents: A plant that can be returned to its known-good state at once.

Why the IT playbook is wrong here

Why the confidentiality-first playbook is dangerous on a control network

On a corporate network the priority order is confidentiality, then integrity, then availability: the worst outcome is data stolen, and a system taken briefly offline during a test is an inconvenience. On a control network the order reverses. Availability and safety come first, because the system is running a physical process — a pump, a furnace, a pipeline — and taking it offline unexpectedly can mean lost production, damaged equipment or a hazard to the people around it. A tester who carries IT habits onto an OT network is the single most common way an assessment causes the incident it was meant to prevent.

That reversal changes the method, not just the caution. A routine port scan can knock an older controller off its process. Automated vulnerability tools written for servers can send input a programmable logic controller was never designed to reject. So the assessment is built to gather most of its findings passively — from network captures, configuration exports and documentation — and to reserve active technique for agreed windows on systems that can tolerate it, with the safety systems left strictly untouched. The result is slower and more careful than an IT test, and that is the correct trade when the thing under test is keeping a physical process within safe limits.

The method

Testing a control network without becoming the outage you were hired to prevent

The sequence moves from listening to acting, and each step forward is earned by the last one showing the system tolerates it. The halt branch is not a contingency; it is the default the whole method is built around.

A flow from passive mapping, to testing the IT-OT boundary, to assessing devices in agreed windows, to stopping at demonstrated reachability — with a halt branch that returns control to operations at any sign of instability.

  1. 01

    Map passively first

    Listen to the network and read the documentation and configuration exports before sending a single active packet to a control device.

  2. 02

    Test the IT-OT boundary

    The segmentation between the corporate network and the plant is where most reachable risk sits and where testing is least likely to disturb a process.

  3. 03

    Assess devices in agreed windows

    PLCs, HMIs and controllers examined during a scheduled window, with an engineer present and a rollback confirmed for anything that changes state.

  4. 04

    Stop at demonstrated reachability

    Show that a control could be reached or a command issued, without actually issuing one that changes plant state — the proof stops before the physical effect.

Always available

Halt and hand back

Any sign of instability stops the test immediately and returns control to the operations team. On a control network the safe default is always to stop, because the cost of an outage outweighs the value of one more finding.

Our role

What we do in the plant and what your operations team signs off

  1. 01

    We adapt the method to the process

    The testing is shaped by what the plant can tolerate, not by a standard checklist — passive where a device is fragile, active only where the process and the schedule allow, and never against the safety systems.

  2. 02

    We come to the site where it is needed

    Part of this work cannot be done over a remote connection: reaching an air-gapped segment, inspecting a device physically, or being in the room with the engineer who can stop the process all require our people on the plant floor.

  3. 03

    Your operations team holds the stop

    Every active step happens with your engineer present and authorised to halt it, because the person who runs the process every day is the one best placed to see instability before an instrument does.

When the plant can accommodate active testing, which systems are too critical to touch, and who holds the authority to stop are decisions for your operations leadership; we bring the offensive-security method, adapt it to those constraints, and treat the safe running of the plant as the boundary the whole engagement respects.

Common questions

What buyers ask

Can you test our OT network without stopping production?+

For most of the assessment, yes, because it is built to be passive: much of what we need comes from listening to the network, reading configuration exports and reviewing documentation, none of which touches a running process. The steps that carry real risk — active probing of a controller, testing that could alter a device state — are confined to agreed windows, run with your engineer present and able to halt them, and kept away from safety systems entirely. Where a step cannot be made safe against live production, it is scheduled into a maintenance window rather than forced, and the report is explicit about what was tested actively and what was not.

Why does some of this work require your engineers on-site?+

Because parts of a control environment are deliberately unreachable from anywhere else. An air-gapped segment has no remote path by design, some devices can only be inspected or connected to physically, and active testing on a live process needs someone in the room with the operations engineer who can stop it the instant something looks wrong. A remote assessment can cover the IT-OT boundary and the network layer, but it cannot safely exercise the devices themselves, and pretending otherwise is how remote OT testing either misses the real risk or takes an unacceptable one. Where the work needs to be on-site, we say so, rather than implying a local capability we deliver from a distance.

Do you test the safety instrumented systems directly?+

Not with active technique. Safety instrumented systems exist to bring a process to a safe state when something goes wrong, and interfering with one during a test is the category of action most likely to cause the harm the whole engagement is meant to avoid. We assess how they are segmented, how they are reachable and how they are protected from tampering — all of which can be established without sending active input to the safety logic itself. If a safety system is found to be reachable in a way it should not be, that is a finding in its own right, and it is reported as one rather than demonstrated by exercising the system.

Next step

Plan a safe OT test window

A scoped OT assessment with agreed windows, excluded safety systems, an on-site plan where needed, and a named engineer who can halt the test the moment a process shows instability.