How AI helps make buildings even smarter

How AI helps make buildings even smarter
Smart buildings have become a much-discussed topic. Users are now looking for buildings that are smarter, more efficient and more energy-saving. Increasingly, artificial intelligence can help in this regard.
That was the point raised by BuildingIQ, which provides energy- and operational efficiency-related services and solutions primarily in the U.S., Australia, Asia and the U.K. According to Steve Nguyen, VP of Product and Marketing at BuildingIQ, their services and solutions are built on top of an AI-driven cloud platform called 5i and are delivered in the form of SaaS subscriptions, project/program management, or on-site services.
According to Nguyen, their solution stands out by way of using artificial intelligence. “Our product is unique in that it combines AI (machine learning) models for the thermal mass of a building, weather forecasts, look-forward pricing, and real-time feedback to create a closed loop of 24/7 energy-optimized control in advance of demand. In other words, our system conditions the environment to reduce energy usage before the demand occurs,” he said. “We leverage this advanced, self-learning model across most of our services to help facility teams do more with the budgets that they have, drive down costs, and identify operational errors before they become a problem. The whole data-rich building environment is monitored 24/7 by a network operations center that acts as the first line of defense in identifying, analyzing, and diagnosing energy-related anomalies. Competitors do only pieces and parts of this.”

Actionable insights from data

AI has become more and more effective and feasible thanks to the availability of data that can be used for meaningful analysis. Nguyen agrees with this notion. “The future holds for more actionable insights from non-correlated data; for instance, the impact of the incident of light on the perception of comfort as it pertains to the actual expenditure on electricity to maintain comfort. There’s lots of work to do, and much of it is driven by data science and machine learning,” he said.
Since the solution is cloud-based, there’s minimal hardware requirements. “In some situations, such as working with utilities, there are no hardware requirements as we get meter data from the utility. In others, we install a lightweight appliance that includes at least a digital meter for the whole building,” Nguyen said. “Within the appliance we typically have software to connect to the existing BMS, protocol translators, and a control translation layer that interprets the outputs of our optimization models into commands that the BMS understands.”
Nguyen expressed optimism for the solution, amid growing demands for buildings that are smart and efficient. “Energy savings is often the tip of the spear and gets us in the door. However, operational optimization and oversight is really the thing that keeps customers happy,” he said. “Comfort has become increasingly important, and for some of our clients, already the most important factor. There is, as you would imagine, often a direct correlation between energy spent on heating and cooling, and occupant comfort. Our system uses comfort ‘bands’ as a parameter so no matter how we’re manipulating zone temperatures, we maintain comfort,” he said.
Finally, the company also attaches great importance to cybersecurity. “Our system is first and foremost dial-out/respond rather than dial-in. We do not maintain a persistent connection to the building. We only speak to the BMS on the operational technology side of the building network. On the operational side, we only access cooling equipment while indirectly impacting heating (no direct control). We (typically) control three pieces of data – supply air temperature, supply air pressure, and zone temperature set point,” Nguyen said. “Architecturally, we communicate with the building using ‘bank’ level security including secure protocols/encryption such as SSL 3.0 / TLS 1.1, AES 256, Public / Private Key, RSA 4096. Client data is virtualized in sandboxed environments. Our cloud is hosted in Rackspace’s secure environment.”

Product Adopted:
Share to:
Comments ( 0 )