How Wertek AI works for buildings (BMS/BACnet)

How it works (concept)

Wertek for buildings is an independent visibility layer on top of the building automation system (BMS) the building already has — Metasys, Niagara, Desigo, Tracer and similar. It does not replace the BMS or write to it: it observes it in read-only mode to answer what the legacy system's screens don't — how many pieces of equipment are actually instrumented, which points deliver reliable data, and what requires attention.

The entry product is the visibility audit: a per-project service run by Wertek or an integration partner, with an on-site visit, a passive capture over the building's BACnet network, and an executive findings report as the deliverable. Unlike the platform usage guides, here the "how you use it" is mainly about how you contract and prepare the audit, and what to do with the result.

Important, so you don't over-expect: continuous subscription monitoring and HVAC control retrofit are in development ROADMAP — they are not contractable as a standard product today; what is contractable is the audit and, afterward, connecting equipment on a per-project basis. What the solution solves and for whom: Wertek for buildings.

First 15 minutes

The audit starts with a scoping call — 15 minutes is enough to get it booked if you have this information on hand:

1. What systems the building has: BMS and supervisor brand, approximate number of controllers, and whether there are chillers, air handlers or meters you'd like to include.
2. Who will grant access to the BMS network: the capture is done by connecting to the network where the BMS lives (typically the supervisor segment). Coordinate the connection point and permissions with your IT team or with the integrator that administers the system.
3. Who accompanies the visit: the building operations lead (or the HVAC integrator) must be present — they know the equipment, help interpret point names and validate findings live.
4. Prior documentation, if any: drawings, an equipment list or the original integrator's inventory speed up the analysis (they are not required).

With the scope defined, Wertek or the integration partner quotes per site (based on size and number of systems) and the visit is scheduled.

Typical step-by-step flow

1. Scope and quote: it is defined which systems are audited and the project is quoted per site.
2. On-site visit: the auditor connects to the BMS network accompanied by the responsible lead. No software is installed on the corporate network and nothing is written to the BMS.
3. Passive capture: discovering devices and points is a matter of seconds to minutes — in the published case of a Class A corporate tower in Monterrey, a 29-second passive capture discovered 57 controllers and 1,161 points behind a single network supervisor (full case).
4. Analysis: what is captured is classified with standard equipment profiles (air handlers, VAV boxes, chillers, rooftop units, fan coils) and the quality of each point is evaluated — which deliver usable data and which return broken values or values outside their physical range.
5. Executive findings report: the central deliverable — an independent document with the complete inventory of devices and points, quantified data quality, the prioritized operational findings in business language (equipment drifting, zones off setpoint, broken sensors feeding automated decisions) and a 30/60/90-day action plan.
6. Findings meeting: the report is reviewed with the responsible lead and it is decided what to fix and what to connect to the platform as the next step.

Common errors and solutions

• "No devices are discovered during the capture" — it is almost always network segmentation: the connection point is not on the same segment/VLAN as the controllers, or a firewall blocks the traffic. Coordinate access to the correct segment with IT before the visit; it is the preparation that saves the most time.
• "The points have cryptic names" — that is normal in legacy BMS (original integrator labels). The audit classifies them with standard equipment profiles, and having the responsible lead on site helps map the ones the naming doesn't reveal.
• "The report shows points with impossible values" — that is not an audit error: it is a finding. Broken or misconfigured sensors report values outside their physical range (in the Monterrey case, an outdoor temperature of 273,708 °C feeding automated decisions). These are among the most valuable findings because the BMS shows them without questioning them.
• "A piece of equipment connected to the platform stopped reporting" — for meters or chillers already integrated on a project basis, check the gateway's power and connectivity on site (internet or data SIM) before escalating to support.

What's next

• Connect what matters, per project: from the report's inventory, energy meters and equipment such as chillers, air handlers and pumps are integrated into the platform — the building's consumption and power quality are managed with the energy guide.
• Close the findings with evidence: the 30/60/90 plan is executed as work orders in the CMMS (guide), with photos and per-equipment traceability.
• Next layer ROADMAP: continuous subscription visibility (permanent BMS monitoring with drift alerts) and HVAC control retrofit are in development; ask about availability when contracting the audit.
• More context on the vertical: commercial buildings.