Machine learning and artificial intelligence is finding increasing application in weather forecasting in an effort to improve what is traditionally a notoriously inexact science.
Utilities are an obvious beneficiary of such advances and some use cases being demonstrated in day to day operations include outage prediction and field worker mobilisation. With weather dependent renewables a growing contributor to the energy mix, other key use cases are renewables dispatch and load management.
One area that has received little attention so far, however, is building energy management and this has now come under investigation by Cornell University researchers with early results demonstrating significant energy efficiency potential.
Machine learning based weather forecasting relies on data on forecasts and actual weather conditions, the longer the datasets the better and with new data continually being added in the goal of improving the forecasting.
The approach taken by the Cornell team, lead by Professor in Energy Systems Engineering, Fengqi You, was to combine the weather forecasting model with a building model taking into account characteristics including the size and shape of rooms, the construction materials, the location of sensors and the position of the windows.
These were then encapsulated into a smart control system to manage, as a demonstrator, heating and cooling systems retrofitted in a 90-year-old building on Cornell’s campus, achieving a reduction in energy usage by up to 10%.
“If the building itself could be ‘smart’ enough to know the weather conditions, or at least somehow understand a little bit more about the weather forecasting information, it could make better adjustments to automatically control its heating and cooling systems to save energy and make occupants more comfortable,” says You.
“For instance, if I know the sun is going to come up very soon, it’s going to be warm, then I probably don’t need to heat the house so much. If I know a storm is coming tonight, then I try to heat up a little bit so I can maintain a comfortable level.”
The Cornell model can detect uncertainty not just in temperature but in precipitation, sunlight and differences in conditions by location.
While You hasn’t indicated how or where the model might be advanced or commercialised, he suggests that in addition to building control systems, other use cases could include irrigation control in agriculture and more efficient indoor environmental control in the vertical farms and plant factories that are emerging in large cities.
Buildings are a key target for energy efficiency and management measures, whether to integrate distributed resources such as solar or storage or to optimise heating or cooling according to time of day, season or varying occupancies.
According to a report from Navigant Research, while the intelligent buildings market has been growing over the last 20-plus years, it is today’s technology innovations associated with the Internet of Things, cloud and edge computing and cybersecurity that have culminated as an inflection point in the market. The result is a broader set of potential intelligent buildings customers exploring investment opportunities and installing solutions.
Today the proposition for investing in intelligent building technologies for energy management is only part of the conversation, according to Navigant. Customers are looking for solutions that translate a complete data profile of their facilities, systems and operations into business metrics around cost, experience and sustainability. The result is an evolving technology landscape with a convergence of building energy management and building management solutions.
“The path forward for owners looking to transform their commercial facilities into intelligent building platforms requires a shift in strategy and processes, investment in technologies and services, and understanding of opportunities that can result from digital transformation.”
As an indicator of the potential, Navigant projects growth of intelligent building solutions in revenue terms of almost five-fold over the next decade.