
FNB · Food & Beverage
Field note · Food & Beverage · Pharmaceuticals
Why “passive” is the only safe first step on a running line
A plant has one question about any work on its network, and it asks it before anything else: will the line keep running?
In a food or beverage plant, a filler that stops means spoiled product and a wasted clean-in-place cycle. In a pharmaceutical plant, a qualified system that stops means a deviation, an investigation, and possibly a lost batch. The scope, the report and the findings all come second to that question, and they should.
What a scan does to a controller
An active scan is an office tool. It sends a packet to every address in a range, asks what's there, then goes back to each one that answered and tries its ports to see what software is listening. A laptop or a server shrugs this off; it's built to receive unexpected traffic all day and ignore what it doesn't want.
A controller isn't built that way. A PLC commissioned ten or more years ago runs a small network stack with a small buffer, designed to talk to one HMI and one engineering workstation in a protocol it expects. A burst of unfamiliar packets, or one malformed packet, can fill that buffer, trip a watchdog, or fault the processor. When the processor faults, the line stops. Older drives, protocol gateways and some instruments behave the same way. NIST SP 800-82 Rev. 3 says so in general terms: tools built for IT networks can disrupt operational equipment, and using them takes planning and care.
In a regulated plant there's a second cost. A scan that reaches a validated controller is an undocumented interaction with a qualified system, whether or not anything breaks. Quality teams have a word for that, and nobody wants it on a batch record.
A scan that stops the line has already cost more than most of what it was looking for.

Listening instead of asking
A passive assessment works from a copy of the traffic. The plant's switch is told to copy packets from chosen ports onto one spare port, a mirror port, and a capture device on that port records what passes. Nothing is sent into the network. From the controllers' side, nothing has happened at all.
The record is richer than it sounds. In the course of ordinary work, every controller, HMI, historian and workstation on the mirrored segment gives itself away: its address, its protocol, who it polls and who polls it, how often, and what crosses into the office or out of the plant. Many industrial protocols carry vendor and model in their normal chatter; some carry firmware too. A few days of capture builds an asset list and a map of who talks to whom.
Its limits, plainly
A method that oversells itself is a method a plant stops trusting, so here are the limits.
It only sees what talks. A device powered on but silent during the window is invisible: a vendor modem that only dials when the vendor calls, a workstation switched off that week, a spare controller still patched in.
It only sees what passes the mirrored ports. An unmanaged switch under a machine hides the conversations inside it.
It infers rather than asks. Some fields stay blank, and the blanks get named as blanks, never filled with a guess.
It shows what's reachable, not what's exploitable. Whether a weakness could actually be used is a different question, answered later and only with authorisation.
And a mirror port is a change to a live switch. Someone logs into a production switch and alters its configuration. On a busy switch, copying many ports onto one can drop packets from the copy, which spoils the capture rather than production; on some older or smaller switches, a configuration change has a real cost in processor time. Where a switch can't be touched at all, the alternative is a network tap, which interrupts the link for a moment as it goes in. That makes it a maintenance-window question.
So passive is lower risk by a wide margin, with a small, known, written-down risk: one switch at a time, done by or beside the plant's own network person, at a time the production plan allows, with the old configuration kept so the change can be undone in a minute.

The order of a passive-first job
Paper first. The drawings, however old. The production plan for the coming weeks. The names of whoever owns each switch. These decide where it's safe to capture, and when. Written authorisation names the switches, ports and window before any cabinet opens. In a pharmaceutical plant this is also when quality confirms that a mirror on a network switch is a documented network change, not a change to a validated system, and records it that way.
Capture at the boundaries. Where lines meet the supervisory layer, where the plant meets the office, where utilities join. A few ports there see the most. The capture runs long enough to cover a full shift pattern and at least one changeover or batch, because recipe traffic, lab transfers and vendor check-ins each keep their own schedule. Meanwhile someone walks the cabinets, because the walk-down finds what the capture can't.
Then the report, in writing. The asset list with its blanks marked, the conversation map, the crossings in and out ranked by what each would allow, and a note on every remote link. Only then, and only if a blank truly needs filling, does anything active come into view: a single query to a single device, in a maintenance window, with written authorisation and a named person beside the work. On most sites like this, that step is never needed.
Uptime is the point
Keeping the line running is the reason to do the assessment at all. The reason to know what's on the network, and what reaches it from outside, is to keep the filler filling and the reactor at temperature. A method that risks either in order to find out has its priorities backwards.
Passive-first keeps the work inside the plant's own rules. No controller is touched, so no change control opens against a qualified system. No line stops, so no product is lost to the assessment. The result arrives in writing, and the plant decides what to do with it at its own pace. ISA/IEC 62443-3-2 asks a plant to know its zones and conduits before it assesses risk; listening is how a plant learns them without paying in downtime. No shutdown. No lost production.
Sources
New to the vocabulary? Learn OT covers it from zero, free.
From the lab


