When an real estate/building owner compares software for a building, the subscription price is the number on the page. The cost of connecting that software to the heating and ventilation controls already on site usually isn’t. And under the way most building software is sold today, that cost comes back with every new application.
Berkeley Lab’s analysis of the US Smart Energy Analytics Campaign (2016–2020) gives a sense of scale. Fault detection and diagnostics tools had a median base cost of $13,000 per building for software and installation, plus $3,500 a year in recurring software costs. In-house staff time came on top of that, at a median of eight hours per building every month.
The usual advice is to give integration its own line in the budget. Tom Novotny, CEO of Finnish company Nobody Engineering, which through its infrastructure lets new software read and write data in the legacy HVAC controls already running in buildings, thinks that only goes halfway. “A separate budget makes that cost visible, but it doesn’t make it go away,” he says.
Owners Find Some Costs Only After Signing
Asked which expenses surface late, Novotny starts with integration itself: work that was underestimated, or scoped before anyone had the full picture of the building. Then come site surveys, support after handover, and the downtime when an integration fails.
Much of this traces back to the installed equipment. Older sites often run several generations of controllers, each with its own interface and its own names for the same measurement. Even systems on BACnet, an open standard for building automation, still have to be configured for each building, according to the BACnet Committee. A temperature reading with no location or unit is of little use to software, and a value refreshed every 15 minutes can’t serve a task that needs one every minute.
Access takes people as well as wiring. A controls technician may have to open a data pathway while the owner’s IT team approves network settings, and Department of Energy guidance recommends bringing network and firewall administrators in early. If the software is allowed to change settings, someone also has to decide whose problem it is when a new setpoint leaves a conference room at 81°F.
Every New Application Restarts the Bill
In today’s model, Novotny says, owners often face a separate integration budget because many software products arrive with their own integration projects. The trouble starts with the second purchase. An owner who adds an analytics tool after an optimization service can end up paying to survey, connect, and test the same controllers again.
Nobody Engineering was set up to remove that line item, he says. The building is connected once at no integration cost, for a small monthly service fee, and any software can use the connection.
The repeat cost weighs even more on software providers, in his view. If every building needs its own integration, a provider has to choose which projects are worth taking and work out how to grow from there. “For them, the question is whether buildings 2, 10 and 100 connect the same way,” he says.
Replacing Working Controls Is the Expensive Option
Replacement makes sense, Novotny says, when the equipment is at the end of its life or failing beyond repair. Until then, an owner has three choices. Doing nothing costs the energy savings the building could have had. Replacing working controls costs time and capital. Connecting them costs a fraction of that, the savings start right away, and the next piece of software uses the same connection.
A retail property of about 15,000 square meters in Stockholm shows the numbers. In Nobody Engineering’s published case with Sally R, a Swedish HVAC optimization company, the building ran an aging Trend Controls system. Replacing it would have cost more than €50,000 per automation cabinet. The controls stayed in place, an edge device established a bi-directional connection with Sally R’s software, and it just took days. Sally R reports about 30% lower electricity use over one year, a figure for the whole deployment rather than the connection alone.
A Pilot Should Show the Portfolio Price
Owners often treat a pilot as a test of the software. Under a project-based model, it says little about the next building, which will have different controllers, point lists, and contractors, and a different integration quote.
Novotny says his company charges the same price per building, so the pilot price is the portfolio price, and the pilot shows the owner how a rollout will run. The team has built a connectivity library to over 75 control systems, he says, and a model it hasn’t connected before “adds a few steps, sometimes a bit of extra time, but no extra cost.”
Whichever model an owner chooses, one question is worth asking before signing: will the next application need its own integration project, or can it use the connection that is already there? The answer decides whether integration is paid for once or turns into a budget line that comes back with every purchase.
Blog recieved via e-mail





























