How Can Legacy Systems Be Integrated Without Creating a Second Silo?

Existing buildings, parks and industrial facilities often already operate BMS, security, energy, access-control, lighting, equipment-monitoring and specialist control systems. Integration should help answer a defined cross-system operating question, not create another dashboard that operators must manage in parallel.

Preserve the native control responsibility

Where an existing system is safe, supportable and able to provide trustworthy data, keep it responsible for its original control function. The integration layer should read, normalise and relate authorised information without bypassing native interlocks, protection logic or diagnostic tools.

Inventory the estate before selecting connectors

Record the system owner, hardware and software version, interface, point type, engineering unit, timestamp source, read/write authority, network boundary, retention behaviour and known issue. Similar labels can carry different meanings, while different labels may identify the same asset. Treat names as evidence to verify, not as a ready-made data model.

Start with a read-only operating pattern

Connect and validate: compare representative points with the native interface and field condition, including units, direction, time, quality state and outage recovery.
Preserve identity: keep the source identifier while adding the asset, location, source system and data-quality context needed by operators.
Integrate the workflow: route selected events to a named owner and retain acknowledgement, investigation, action and closure evidence.
Expand by proven pattern: add systems, sites or use cases only after the first scope has stable data and clear ownership.

Treat write-back as a separate risk decision

A technically writable interface is not an approval to control equipment. Define an approved command list, permissions, interlocks, failure behaviour, test evidence and rollback plan for each write-back use case.

When integration should not be the first step

Remediate or replace the underlying component first when a critical controller is unsafe, unsupported or unable to provide trustworthy data. Pause broad integration when asset ownership, maintenance practice or network-security responsibility is unresolved. If the only value is a common screen with no response workflow, the project may create a second silo even when the connections work.

Webuild Tech's Physical AI platform and smart operations scenario can support governed connectivity, asset context, shared visibility and workflow coordination. Compatibility, interfaces and write-back scope must be verified system by system.

For a deeper procurement and implementation framework, read How to Integrate Legacy Building Systems Without Creating a Second Operations Silo.