Posted on June 1, 2018 by staff

Could smart buildings become energy suppliers?


New artificial intelligence technology could make buildings smart enough to become their own energy supplier – and even sell back what they don’t use.

That’s the view of Vijay Natarajan, co-founder and marketing director of AI-powered energy analytics business Qbots, which is based at Manchester Science Park.

“We enable buildings to actively participate in the energy markets, rather than just passively receive energy bills,” Natarajan explained.

This way of looking at energy bills is still relatively new. Most buildings purchase electricity from a supplier, which Natarajan says costs twice as much as the energy itself.

“The rest of it is the other levies and charges,” he told BusinessCloud.

“It is based on when you use energy, including peak time charges and winter-time charges.”

But when a building is capable of reducing its energy demand at certain times by utilising flexibility in its systems or through energy storage, it can then sell energy back to the grid at a premium rate.

This concept is a fairly new one. There are not yet any regulations around such energy trading but Natarajan, who has a background in mechanical engineering, believes that a combination of Qbots’ technology, solar power and storage batteries can bring it into reality.

“People are already working on peer-to-peer energy trading,” he explained.

“We see ourselves as the enabler. Once we have the Qbots technology in a building, it is ready to talk to markets and talk to other buildings.”

The company makes use of AI to precisely assess and predict energy consumption in real time.

This allows the building to store off-peak energy and to invest in solar power tech without relying on peak-time electricity from their supplier.

“If you want to predict the energy demand of a building accurately, you can make a basic prediction with a smart meter data, last year’s data and a record of the weather.”

But he says the process of developing a prediction model for day-to-day control is time-consuming and, if anything changes from the previous year, the predictions are no longer accurate.

“Someone could put up some new furniture – even that could affect everything,” he said.

“If you did real modelling of the system it would take ages and more importantly it would never be right.”

That’s where Qbots AI comes in. “We just get the data, there is no building model,” Natarajan explained.

“We want to create something that can be used for control. We don’t just want a prediction for the sake of prediction, but instead to be able to use that prediction to make changes.”

The company calls its technology a ‘decision support system’ for energy consumption, providing building managers with energy saving schedules that no human could design.

Though not required, the software works with a building’s own power supply too – both from green energy, like solar panels, and cheap energy stored in large, on-site batteries.

Again, without accurate data to know when the best time is to discharge these batteries, the investment does not pay off.

“If you have a battery everything is more flexible, you can charge it and discharge it when it is best for you. You can avoid peak time charges by charging the battery at night time,” said Natarajan.

As buildings become better at creating their own energy, and more efficient at saving it, they remove themselves from the dependent relationship with their supplier.

“You can become energy independent, and every building becomes a sort of power station. That is the approach,” he said.

These ‘smart’ commercial buildings can then consider selling surplus energy back to the provider or to a nearby connected smart building.

“It’s like Airbnb, energy can become like that,” he said.

“You have the facilities, some time to spare and the enabling technology to be part of the sharing economy.”

Qbots is currently trying to find funding with EDMI to bring the same intelligence to small businesses and homes.

EDMI creates smart metering hardware and are involved in the government roll-out of smart meters into homes and small businesses.

Qbots hope to bring their brains to these devices, enabling people to better manage their own energy consumption and join the energy market.

“After it is installed you’ll be able to not only save but to interact with the market and earn revenues.

“There are lots of new energy companies in the last few years, so things are changing.”

The company is currently gearing up to demonstrate its technology as part of the CityVerve project, which aims to demonstrate the future of smart cities.