According to IBM researcher Arun Vishwanath, disruption is coming to the building controls industry in the form of artificial intelligence and machine learning that could eventually enable self-driving buildings.
Vishwanath was the keynote speaker at AIRAH’s inaugural Big Data and Analytics Forum in Sydney last week. The event saw 120 attendees – including analytics consultants, BMS operators, building services contractors, facilities managers and property owners – discussing the latest developments in IT-assisted improvements to buildings’ operational efficiency.
Vishwanath unveiled the new IBM system called BEACH (Building Energy Analytics for Cooling and Heating). He also presented a case study of how it has been successfully implemented for a large office building in Townsville.
BEACH helped optimise the staging and operation of the building’s chillers (which alone represent 40 per cent of the building’s total energy consumption) and HVAC system, while maintaining occupant comfort. Since the system was implemented in December 2018, it has saved 15–20 per cent of cooling energy.
Perhaps more significantly, BEACH operates on the IBM cloud platform. The system uses secure RESTful APIs (application programming interfaces) alongside the Project Haystack open-source IoT initiative to autonomously adjust cooling start times and zone set-point temperatures, irrespective of the BMS.
It represents – as predicted by some industry observers – the entry of a traditionally technology-focused company into the HVAC&R space. Speaking after his presentation, Vishwanath predicted that a major disruption is coming to energy management in buildings.
“IBM is not a building management company,” he said. “I’m controlling a third-party BMS, and I don’t even know who the BMS vendor is. The APIs make it happen.
“So you don’t have to necessarily tie your solution to one BMS vendor or the other. The emerging standards – open-source standards like Haystack and Brick – make it possible for third-party analytics to run on any third-party BMS. They don’t need to know each other. They don’t need to be shaking hands.
“Maybe in the next five or 10 years a lot of start-ups will come up with fantastic artificial-intelligence machine-learning (AI ML) algorithms and control a BMS provided by another company. You’re going to have buildings that are completely self-driving, like self-driving cars. And you don’t need dedicated BMS sitting in every single site. You can have the AI ML models running it and pretty much have a simple local controller that responds to your decisions.”
During the presentation, Vishwanath demonstrated remote monitoring and control of the Townsville site.
“The controls companies should pull up their socks,” he said. “The data is there. I have shown you what I can do as an IBM-er, not at all in the BMS system, controlling the building provided by a third-party company. If Townsville likes what we do, the question they will ask is, why do I need to keep paying the BMS every year, when IBM can do a better job. That is a disruption waiting to happen.”