Systems About 3 min read

BMS and SCADA integration

Define the data exchanged with building or industrial systems. Check ownership and approval for any controls exposed through the integration.

Sources and scopeSource record 25 August 2026

Technical source record: 25 August 2026. Check the linked documentation for current product requirements.

Defensive architecture only; no register/object writes, live commands, controller logic, safety function, process setting, commissioning procedure, or site specific cause and effect.

Verification and testing

Overview#

Building management and industrial control systems affect real processes: HVAC, power, water, smoke control, lifts, machinery, environmental conditions, and other services. Security integration should default to constrained, observable exchange. A common dashboard never becomes engineering or safety authority by presentation alone.

Boundary architecture#

text
native controller/process <-> native supervisory system and approved safety logic
              |
        bounded gateway/DMZ or broker
              |
     read-only events/state/trends -> security correlation
     separately approved semantic request <- constrained workflow

Prefer native, supported integration interfaces and a broker/gateway over direct controller access. Keep management and control networks separated according to risk and necessary flows. Avoid dual homed general purpose hosts that bypass the boundary design.

Authority map#

Information/action Authority
process value and quality field/controller/native supervisory path
safety interlock and protective action approved native safety/control design
engineering units, scaling, alarms, limits native engineering configuration
security event/case source security system and operations workflow
cross domain correlation integration platform, clearly marked derived
command or set point change separately authorised control workflow; never inferred from dashboard access

Security staff visibility doesn't imply permission or competence to operate plant. OT staff visibility of a security event doesn't grant access control, alarm reset, or surveillance authority.

Point and event semantics#

For every exchanged point retain native object/address reference, human meaning, source, units, scale, range, enumerations, read/write capability, quality/status, sample/update mode, source time, receive time, maximum age, deadband, alarm semantics, and mapping version.

Distinguish commanded value, controller accepted value, feedback/process value, alarm state, acknowledgement, inhibit/override, maintenance, and communication failure. A successful protocol response or write acknowledgement doesn't prove plant movement or desired physical outcome.

When mapping BACnet, Modbus, OPC UA, KNX, DNP3, or other protocols, use their own security and semantic limits. See Building and industrial protocols. Protocol normalisation must not discard native quality, type, units, or security context.

Approved use patterns#

Lower risk examples include receiving equipment fault, environmental, power, or room condition observations for correlation; publishing a constrained security status to an authorised native consumer; and associating incidents with approved trend data.

Higher risk patterns requiring explicit engineering, safety, and operational approval include changing set points or modes, suppressing alarms, starting/stopping equipment, smoke control interaction, lift commands, door/gate release, power switching, and automated cross domain cause and effect. This page provides no implementation procedure for them.

Cybersecurity and resilience#

  • Inventory assets, firmware/software, protocols, data flows, owners, remote access, and unsupported dependencies.
  • Authenticate users and workloads; restrict service identities to exact objects/operations.
  • Make read only technically enforced where only observation is intended.
  • Use controlled remote access with approval, monitoring, time bounds, and revocation.
  • Protect engineering configuration, logic, backups, keys/certificates, and update paths.
  • Monitor gateway health, stale points, rejected/unauthorised requests, configuration drift, time error, and unexpected write capability.
  • Design for safe native operation if the security platform, gateway, identity service, WAN/cloud, or time source fails.

NIST SP 800-82 Rev. 3 describes OT security while recognizing performance, reliability, and safety requirements; its scope includes building automation and physical access systems. Apply it with the site's engineering and safety obligations.

Change and validation boundary#

Maintain a signed off data flow and point list; owner for every mapping; approved direction; configuration/version evidence; rollback; maintenance window; and post change review. Validate in an authorised environment and coordinated site process, including stale data, loss, partial failure, recovery, and native fallback.

Keep exploratory validation disconnected from live controllers, plant, fire/smoke control, lifts, locks, gates, relays, and safety functions. Coordinate any physical acceptance activity through the site change and safety process.

Questions before connection#

  • Is the business purpose achievable with one way/read only data?
  • Which system is authoritative for each value and decision?
  • Are units, quality, source time, stale state, and override visible end to end?
  • What common dependencies or new remote paths does the gateway create?
  • Can either side continue safely and securely when the integration is wrong or unavailable?
  • Who approves, observes, reverses, and investigates every privileged change?

See Integration platforms and Gates, barriers, and vehicle access.