Rising energy prices and regulatory pressure to limit GHG emissions both push companies to find every possible solution to save power. There has been significant investment in energy, with nearly $1.9 trillion spent in 2018 alone. But improvements have been unfocused, combining elements of sustainability with general power-management principles. Similarly, technological innovations in AI and machine learning are only recently being applied to energy management. So how can a company employ artificial intelligence to direct its energy strategy?
AI in Metering
While it might seem obvious, it is important to remember that management requires information. All facilities must collect data from energy consumption points. Armed with this information, managers can understand when and where facilities use energy, be it power, water, steam, or gas. This can be used to develop an energy consumption reduction strategy. Life sciences in particular can struggle with understanding how their facilities use energy. Since the industry has more in-depth humidity, light, and heat needs, HVAC systems, for example, will work very differently. Weather conditions can create variability, so managers need longitudinal data, as one month might significantly differ from another. Regular data gathering can also help observe trends in energy consumption, so managers can make informed decisions. Creating an energy use profile will also help offset human errors. Imagine an operator using a higher-energy setting on a machine for a one-time event. Should the operator not revert the setting, the machine could be using more energy than necessary in all subsequent shifts. AI offers a number of advantages in energy metering. Firstly, there is less room for operator error. Secondly, the mass of data gathered through metering may be unwieldy or hard to interpret. Artificial intelligence can parse information significantly quicker than a human.
AI does not just work with existing data, it has the potential to create predictive models about future energy use. In a Danish wastewater treatment plant, AI-guided control algorithms were effective at reducing GHG emissions. Simulatainiusly, the predictive model helped cut operation costs, compared to the previously used control method. Other algorithms could also help a company forecast future energy prices. Fed regular data from smart meters, together with the massive amounts of energy price figures, an AI can extrapolate future price variants. Existing AI tools have the ability to create models of a building or facility. The system will develop a plan for the next 24 hours based on a variety of data. For example, it can read environmental data about weather forecasts. Knowing that the next day is likely to be warmer, it can limit the amount of heating earlier to save energy. Similarly, neural networks can also help forecast when machines and infrastructure might need maintenance.
So what can companies do now?
Firstly, understanding the trends in the energy market is vital when considering costs and emissions. It is unlikely that the market will retain stability as prices fluctuate and governments push to limit emissions. Other extreme events also have the potential to disrupt supply and prices. AI has the potential to help a business understand what the future holds, at least regarding prices and short-term consumption. Secondly, companies should make a holistic assessment of where they can implement AI in their operations. This can require a significant amount of digitalization, with its own short-term costs. But the possible gains are enormous. Some solutions on the market boast that they can reduce energy costs by up to 60%. While this figure may only apply in fringe cases, there is growing evidence that AI can help with efficiency.
Lastly, companies need to learn from their peers who have embraced this technology. The Net Zero Food & Beverage Forum: Energy Efficiency and Decarbonisation conference, held from the 16th to 17th of February is an ideal opportunity. Industry leaders and innovators will gather in Berlin to discuss and exchange ideas and experiences with managing energy consumption.