A building operations dashboard on a screen showing connected energy and equipment data

Owners Now Call AI Essential to Running a Building. Most Cannot Actually Use It Yet.

Data Center Demand and Innovation| Electrification Economics at the Property Level

August 04, 2026•5 min read

New research finds that 59% of real estate leaders expect AI to transform how buildings run within three years, but only 36% have actually scaled the technology across their portfolios. That 23-point gap exists because most buildings lack the basic wiring of data, connected meters, sensors, and controls, that AI needs to work. The fix is unglamorous and it comes first: connect the building's data before buying the smart software, or the software has nothing to read.

By Keith Reynolds | Publisher & Editor, ChargedUp!

Home | All Stories

The Gap Between Wanting and Doing

Research from Siemens' Infrastructure Transition Monitor found that 59% of industry leaders expect AI to transform building operations within three years, while only 36% report real progress in scaling AI and digital-twin technology across their buildings, as reported by Propmodo. A digital twin is a live software model of a building fed by its real sensors and meters, used to spot problems and test changes. The 23-point gap between expecting AI to matter and actually using it is the clearest sign yet that AI has shifted from an experiment to an operating requirement, and that most owners are still building the foundation to use it.

Why does an owner care about a technology-adoption statistic? Because the applications on the other side of that gap all target the operating costs that determine net operating income, the money a building keeps after expenses. Predictive maintenance catches a failing chiller before it dies. Automated controls tune heating and cooling to occupancy. Energy management trims the single largest controllable line on many commercial bills. Each targets a recurring cost, and each depends on the same thing: usable, connected data. A fault-detection system cannot flag a struggling chiller if the system watching it is walled off from the meter that would confirm the problem.

Where the Gap Actually Lives

The barrier is rarely the software. It is the building's plumbing. The research identifies fragmented data and disconnected building systems as the specific obstacles slowing progress, per the same Siemens research. This is a portfolio problem, not a single-building one. A team can pilot smart fault detection on one modern building. Doing it across dozens of buildings of different ages, with different equipment and different metering, is a much harder job.

That has a direct effect on capital planning that is easy to miss. If a building lacks the meters, controls, or data setup to support advanced analytics, the first check an owner writes may need to go toward connecting those systems, not toward the AI software itself. Framing digital modernization as buying software, rather than as an infrastructure upgrade that has to come first, is how budgets and timelines get blown. The order of operations is the whole point.

Short-Staffed Teams Make It Urgent

A second pressure makes this practical rather than optional. The real estate services firm CBRE reports that 43% of United States facility teams are understaffed, even as buildings get more connected and generate more data, cited in the research. More connected equipment produces more alerts and more things to check, and a short-staffed team has less time to sort through them. Without automation to filter the noise, more technology just means more work.

This reframes what the automation is for. It is not there to replace skilled staff. It is there to filter the flood of signals, flag the failures that matter, and recommend fixes, so a smaller team can spend its time on judgment calls and vendor decisions instead of sifting alerts by hand. That matters most in buildings that lean on the knowledge of a few long-tenured people, where one departure or one busy stretch can quietly erode how well the building runs.

One Project Shows the Right Order

The Pennsylvania Convention Center in Philadelphia shows what closing the gap looks like in practice. Facing a downturn in convention business during the 2020 pandemic, it committed to a modernization program built around a long-standing partnership with Siemens, as documented by Facility Management. A financing arrangement funded roughly $24 million in smart-infrastructure investment across two phases. The important part is the sequence.

The money first went to meters, sensors, and analysis software that pulled electricity, gas, and water data into one place, creating the visibility needed to find waste. Only then came the automated controls and fault-detection tools that flag equipment problems before they cause downtime, along with automated demand response that cuts power use during expensive peak hours in exchange for lower rates. The center credits the broader effort with cutting energy use by 18, and a separate lighting retrofit cut annual power use by 4.5 million kilowatt-hours. The lesson is the order: data and metering first, smart controls second, savings and certifications as the result rather than the starting point. That ordering is precisely what the research suggests most portfolios have not yet done.

The Planning Test Owners Cannot Defer

The central finding is that connected data and AI have become operating necessities for any portfolio trying to protect income and manage rising costs, not optional upgrades to weigh on a long timeline. The unresolved question for most owners is not whether to invest but in what order. A portfolio with fragmented systems will struggle to catch avoidable energy waste no matter how good its software is. A short-staffed team will miss early warnings without automation to prioritize them. And a building without clean, centralized data will have trouble proving its efficiency to lenders, insurers, or reporting requirements, even when the upgrades are technically underway.

The Pennsylvania Convention Center points to a specific answer: metering and system integration first, automated controls and predictive maintenance second, certifications and durable savings as the outcome rather than the beginning. For any owner still weighing whether to fund the data foundation before the AI tools, that order is the practical test the next planning cycle will have to resolve. The building that connects its data now is the one that can use everything that comes after.

Sources

Keith Reynolds

Back to Blog