William Webster

written by

William Webster

Researcher, Avoncourt Partners GmbH

Culture Blog - Mar 5, 2018

AI will help drop energy consumption

“Before we work on artificial intelligence why don’t we do something about natural stupidity?”

These words by Steve Polyak reflect both the incredible heights to which human intelligence can reach, but also the utter lack of knowledge of our own intellectual and neural capacities. We know much more than we can talk about, yet we catch ourselves daily “doing without thinking,” which is the essence of natural stupidity.

Close of male hand holding brain between fingers

One vital area for humans to improve applied knowledge and overcome natural stupidity is energy consumption.

And one of the greatest instruments for overcoming it is on the cusp of revolutionizing it: artificial intelligence.

People waste energy by leaving lights burning unnecessarily, letting the faucet flow while not in the shower, leaving the car motor running while not needing an engine’s power, to name a few. Corporations have been looking for ways to optimize (and minimize) energy use for decades. Some nations have created habits of frugal private use of energy. The World Bank has provided online resources to help inform about wasteful energy consumption in cities around the world.

Yet knowing does not always equate with doing, aka our “natural stupidity.“

Artificial intelligence start-ups are providing energy corporations with intelligent solutions to the complex problems of energy consumption.

Unlike in other fields, energy demand is always in real time, needed immediately. The blackouts that occur frequently in many 2nd and 3rd world countries are due to a lack of planning for high demand periods. Making short-term load forecasts more accurate in order to adjust supply to meet anticipated demand can deliver enormous savings, reduce waste and emissions, and add to system resilience.

Artificial intelligence can enhance demand and supply prediction, assess the reliability of integrated generation assets, and automate the demand-side response. This would no longer have to be the complex work of engineers and executives, but can be achieved with a single bot. The United Kingdom is, for example, planning to cut 10% in national electricity usage by using deep learning to predict power demand and supply. UK’s National Grid is collaborating with DeepMind, an AI startup bought by Google in 2014, to predict supply and demand variations based on weather-related variables and smart meters as exogenous inputs.

In addition, energy production can be optimized through the help of AI in preventive maintenance, as well as improving the electricity production yield, reducing energy waste, and preventing electricity theft. McKinsey Global Institute reported in July 2017 that some countries see a 20% energy production increase using machine learning and smart sensors to optimize their assets’ yields. By optimizing pricing with time-of-day and dynamic tariffing, and matching producers and consumers in real time, energy corporations have used AI to increase profits by 10-20%. 

In homes, private individuals can utilize artificial intelligence to automate the supplier selection and provide consumption insights. This could result per capita in $10–$30 savings on monthly bills by using machine learning to automatically switch electricity supply deals.

Female home owner checking energy consumption on digital tablet using smart meter technology.

AI-powered technologies can help deliver more efficient designs than were previously achievable by eliminating waste in things like space heaters and radiators. Heat exchangers have already been created using generative design, reaching levels of efficiency that the human mind would need decades to perfect.

All of these developments help curb the human problem of natural stupidity in energy consumption. This is just one area where AI can have a tremendous effect on the planet’s ecosystem, helping countries minimize their CO2 outputs and providing coming generations a promising future.