Best practice

Unmanaged devices on OT networks: find and fix them

How to use an OT asset inventory to find unmanaged devices on industrial networks, decide what to do with each one, and stop new ones appearing.

OTVista article cover: Unmanaged devices on OT networks: find and fix them. Photo: Close-up of control panel indicator lights

Almost every industrial site has devices on its networks that nobody actively manages: a cellular router a vendor installed for remote support, a spare engineering laptop left plugged into a cabinet switch, a small PLC added during a line change, an IP camera on the control network, or a test box that was never removed. We call these unmanaged devices. They are not necessarily malicious, but they have no owner, no patch or backup routine and often no documented purpose, so they are exactly the kind of asset an attacker or an honest mistake can exploit.

The good news is that the tool for dealing with them already exists in most OT security programmes: the asset inventory. This article describes how OTVista researchers would use an inventory to find unmanaged devices, decide what to do with each, and keep new ones from appearing.

What counts as an unmanaged device in OT

A device is unmanaged when one or more of these basic questions has no answer:

  • Who owns it? There is no named person or team responsible for it.
  • Why is it there? Nobody can say which process, function or vendor needs it.
  • Is it maintained? It is not covered by patching, configuration backup, hardening or monitoring routines.
  • Is it recorded? It does not appear in the inventory, drawings or the configuration management records.

Some unmanaged devices are obvious, such as a consumer-grade switch under a desk. Others look legitimate: a historian replica built for a project, or a second network interface on an HMI that bridges two zones. Both kinds matter.

Finding them: compare what you know with what is there

Unmanaged devices show up as differences between sources. The practical method is to build the inventory from records first, then compare it with what the network actually shows.

  1. Start from records. Engineering drawings, PLC and DCS project files, maintenance systems and purchase records give the list of devices that are supposed to exist. Our OT asset inventory guide lists the attributes worth capturing.
  2. Observe the network. Passive monitoring of traffic at agreed points, switch MAC address and ARP tables, DHCP leases and firewall logs show which devices are actually communicating. Any active querying should be agreed with asset owners and vendors first, because some older controllers handle unexpected traffic poorly.
  3. Walk the site. Cabinets, marshalling rooms and remote enclosures reveal cellular modems, serial-to-Ethernet converters and devices that rarely talk on the network.
  4. Reconcile. Each device seen but not recorded is a candidate unmanaged device. Each device recorded but never seen needs checking too: it may be offline, isolated, or already removed.

The comparison is described in more detail on our OT asset visibility page. The joint guidance from CISA and partner agencies, Foundations for OT Cybersecurity: Asset Inventory Guidance for Owners and Operators, gives a useful baseline for the attributes and process.

Triage: decide what to do with each device

Once a device is identified, treat it as a small finding rather than a cleanup chore. For each one, record what it is, where it sits, what it talks to, and then make one of four decisions with the asset owner:

  • Adopt it. The device is needed. Give it an owner, add it to the inventory with full attributes, bring it into patch, backup and hardening routines, and document its purpose and permitted communications.
  • Replace it. The function is needed but the device is unsuitable, for example consumer-grade hardware or an unsupported operating system. Plan a replacement in the next suitable maintenance window and apply compensating controls until then.
  • Isolate it. The device must stay for now but should not have its current reach. Restrict it with firewall rules or move it into an appropriate zone.
  • Remove it. Nobody needs it. Agree removal with operations, check for hidden dependencies first, and record the change.

Prioritise by exposure: devices that provide a path from outside (cellular routers, remote access boxes, dual-homed machines) or that bridge security zones come first. An OT security architecture review is the natural place to assess those paths.

Keeping new unmanaged devices from appearing

Finding devices once is not enough; the same causes will produce new ones. The controls that help most are procedural as much as technical:

  • Change management that includes the inventory. No device is connected, moved or removed without a change record that updates the inventory. Vendor and contractor work is included.
  • Rules for vendor equipment. Remote access devices and test equipment brought by third parties are approved, recorded and removed when the work ends.
  • Periodic reconciliation. Repeat the records-versus-network comparison on a fixed cycle and after major projects or outages.
  • Port and access controls where practical. Disabling unused switch ports and alerting on new MAC addresses can catch additions, but test these carefully in OT, where a blocked port can stop production.
  • Clear ownership. Every zone and every system has a named owner who signs off the inventory for it.

These rules belong in the site's written procedures; our page on OT policies and procedures covers how asset management and change management procedures fit together.

Where to start

If the problem feels large, start with one site or one zone, ideally the one with the most external connectivity. Build the record-based inventory, compare it with a short period of passive observation, triage what you find, and fix the change process that let the devices in. That single cycle usually shows how big the problem really is and which controls will matter most across the rest of the estate.

Cover photo: iSawRed on Unsplash, used under the Unsplash License.

Topics: unmanaged devices OT, OT asset inventory, OT asset discovery, shadow OT, OT asset management

Frequently asked questions

What is an unmanaged device on an OT network?

It is a device with no named owner, no documented purpose, or no maintenance routine such as patching, configuration backup or monitoring. Common examples are vendor cellular routers, spare laptops, unrecorded PLCs and devices left behind after projects.

How can an asset inventory help manage unmanaged devices?

Build the inventory from records first, then compare it with what network observation and site walkdowns show. Devices seen but not recorded are candidate unmanaged devices; each is then adopted, replaced, isolated or removed with the asset owner.

Is active scanning safe for finding unmanaged OT devices?

Only when agreed with asset owners and vendors, carefully scoped and scheduled. Some older controllers react badly to unexpected traffic, so passive observation, switch tables and site walkdowns are the usual starting points in live plants.

How do you stop new unmanaged devices appearing?

Tie change management to the inventory, set rules for vendor and contractor equipment, reconcile records with the network on a fixed cycle, and give every zone a named owner. Port controls can help but must be tested carefully in OT.

See OTVista in action

Join the research preview or talk to us about your OT security priorities.

Join the research