Hospital corridor with an imaging suite door and a network cabinet in a plant room

HLT · Private Healthcare

Field note · Private Healthcare · Buildings & Facilities

Patch Tuesday doesn’t exist on the plant floor

Range Lab2026-11-036 min read

In the office, patching has a beat. Updates land on a schedule, laptops restart overnight, and a good IT team is judged by how few days pass between a fix being published and a fix being installed.

On the plant floor, that number can be measured in years. That isn't neglect. The reasons are good ones.

A patch can stop the process

Controllers are tested as a set. PLC firmware is validated by the vendor against a particular HMI, a particular engineering tool and a particular set of function blocks. A version step can change timing, drop a connection, or leave a program that no longer loads. The vendor validates its fix, then the plant validates it again, on a test rig if there is one and on the line if there isn't.

Medical devices are certified as supplied. The computer inside a scanner, an analyser or an infusion pump is part of the regulated device. The manufacturer decides which operating system and updates it runs, not the hospital. Install an unapproved patch on a modality and a hospital can lose its certification and its service contract in one go.

Validated systems need change control. In a pharmaceutical plant, or any site under GMP or something like it, changing a qualified system means a documented request, testing, approval and a record, all before the next batch. That is right for the product. It also makes every patch a project.

So patching happens in maintenance windows: planned shutdowns, a few times a year at most, and only for fixes the vendor has validated.

The useful question is what protects the equipment in the years between fixes.

Protection around the box

A compensating control protects a system from the outside when you can't change the inside. ISA/IEC 62443-3-3 calls it a compensating countermeasure: when a component can't meet a requirement by itself, something else in its zone meets it on its behalf. NIST SP 800-82 Rev. 3 says the same about patching. Where a patch can't go in, other measures carry the risk meanwhile.

Five kinds do most of the work:

  • Segmentation puts the unpatched system in a zone of its own, so trouble elsewhere can't reach it.
  • Allow-listing at the boundary passes a named list of sources, destinations and protocols and drops the rest. An exposed service on an old controller becomes one only its HMI can reach.
  • Brokered remote access means the vendor connects through a jump host the plant controls, opened on request, time-boxed and logged, instead of a VPN box with a tunnel that never closes.
  • Monitoring means a passive capture at the zone boundary, so the plant sees what talks to the device and notices when something new does.
  • Physical controls are the locked cabinet, the disabled USB port, the engineering port that needs a key, and the rule that a service laptop gets checked before it's plugged in.

None of these touches the controller or the scanner. They fit around the certification, the validation and the vendor's warranty, because they change the network and leave the device alone.

Wide view of a hospital plant room with air handling units and switchboards
PHC · 08 / 09 · wide

The computer inside the scanner

Healthcare shows this most sharply. A private hospital runs two plants at once. The building plant: chillers, air handling, medical gas, generators. The clinical plant: imaging, lab analysers, monitoring, infusion. Both share the same patching problem. A scanner's console runs whatever it shipped with until the manufacturer says otherwise, and an air handling unit's controller keeps its commissioning firmware.

They usually share a network with the office, because the hospital grew faster than its network was redesigned. On a flat network the oldest device sets the risk for everything, and a workstation that reads email sits one hop from a modality nobody can patch. The fix is to give imaging its own zone, with DICOM and HL7 allowed across the boundary and nothing else. The device stays exactly as supplied, and the hospital's own systems stop being its neighbours.

Active scanning is a real risk here. A scanner that stalls during a port scan means a patient waiting. An infusion pump dropping off the network is worse. Listening shows what devices talk to, which vendors reach them and which paths cross into the office, without sending them a single packet. On a clinical network the method is passive first and passive only, unless the hospital asks for more in writing.

The BMS that keeps a theatre at pressure

Buildings carry the same problem more quietly. A building management system runs the air handling that keeps an operating theatre at positive pressure, the chillers that keep a Gulf building usable, plus lifts, access control and smoke control. Its controllers run for years without a restart. They speak BACnet, Modbus and KNX, none of which checks who is talking.

A BMS controller is rarely patched: a firmware update means taking a plant room offline, and often only the original contractor knows how. The same compensating controls apply. A BMS zone separate from tenants and the office. The contractor's gateway brokered, not left open. A locked plant room. A capture at the head end, so the facilities team knows which systems share a network and which gateways reach outside.

Close view of a BMS gateway cabinet with DIN-rail controllers and a labelled breaker
BFM · 05 / 09 · detail

Hospital and building meet in the theatre. The BMS holding the pressure shares a cable with the clinical floor more often than anyone would design, and neither side of that cable is getting patched this month.

Two columns, not one

A plant's patch plan needs two columns. The first lists what can be patched, when the next window falls, and which vendor validations are pending. The second lists what can't be patched, for how long, and what stands in for the patch meanwhile. On most sites the second column is longer, and it's the one a plant manager should be able to show.

Sample · illustrative, not a client site

System Patchable Next window Standing in for the patch
Line 2 PLC Yes, vendor-validated firmware only Annual shutdown Own zone, allow-list at boundary, brokered vendor access
CT scanner console No, manufacturer-controlled Not planned Imaging zone, DICOM and HL7 only, no internet, locked console
Theatre AHU controllers Not in practice Contractor visit BMS zone, brokered gateway, locked plant room, capture at head end
Engineering workstation Yes, monthly Monthly No email, USB disabled, kept inside the control zone

Range's part is the first step: the list of what's on the network, what each system talks to, which outside paths reach it, and which of those a patch would never have closed anyway. It comes from listening, so it works on a validated system without a change request and on a scanner without a service call. After that the plan is the plant's own.

New to the vocabulary? Learn OT covers it from zero, free.