Ask the average energy consumer what changes they’d like to see in the energy sector, and it’s unlikely that artificial intelligence (AI) will make the list. Without running a national survey, it’s safe to say that lower energy costs would make up the overwhelmingly majority of responses. Meanwhile businesses more attuned to the challenges facing the industry would perhaps add “greater reliability and efficiency” to their wish list.
No industry is immune to the great changes promised by AI, but the problem is that the technology is often seen as peripheral – interesting, no doubt, but incapable of solving the most urgent challenges that worry the majority of energy users.
That may be true to varying degrees in other sectors, but not in the energy industry. Here, AI is playing an integral part in shaping the way we generate, transmit and use energy. Most importantly, it promises ways to control our energy use (both at macro- and individual levels), improve reliability and keep down the costs of rapidly rising energy bills.
Energy affects us all, whether it’s those who suffer from a lack of electricity, or those in the developed world who want to power their businesses and devices with less damage to the planet (and at lower cost).
It’s no wonder that energy is seen as an exciting new frontier for the technology industry. Most famously, perhaps, is Elon Musk with his vision of a future where we will all drive (or be driven by) high-performance electric vehicles. But many other new businesses are just starting to come online, all of which promise to change our relationship with energy.
Innovations in energy sector technology, such as energy storage battery options and mobile-based thermostat apps, are becoming consumer centric at a speed that has surprised the wider utility industry, as well as the adopters. To meet this ever-increasing rise in demand from their consumers, energy utility companies are integrating these types of innovations into their everyday operations.
In fact, even enterprise customers are progressively starting to monitor, control and optimize energy usage patterns in real time. For example, manufacturers are monitoring energy consuming equipment performance with smart IoT sensors which can predict malfunctions, prevent outages, or even alert when preventive repair is needed. In addition to this, commercial building owners and shopping malls are employing intelligent energy meters to help them gauge consumption patterns and predict their future energy demands upfront.
Companies like Estonia’s Sympower, for example, plan to improve energy responses through demand response aggregation that synchronizes energy supply with real-time demand garnered from the burgeoning array of connected appliances. The Italian start-up DAJIE uses the power of blockchain to enable peer-to-peer sharing, while Greece’s Wapo.io is a virtual facility manager that shares energy data widely via an AI-powered chat bot.
What connects these and hundreds of other new businesses is that they all rely on AI to bring their benefits to a market that is hungry for new solutions to intractable energy problems. Artificial intelligence is crucial for managing and optimizing the incredibly complex infrastructure of power generation (including micro-generation), grids, localities, homes and connected devices. As such, AI isn’t just a toy for start-ups, but should be central to every energy business’ future strategy. In fact, according to our own research, 48 percent of the respondents from the energy, oil, gas and utilities industries agreed that AI is fundamental to their organization’s success. Moreover, an additional 46 percent said that their organization is “building AI into the company” ethos, further reinforcing the need to incorporate elements of AI into the energy industry.
As we know, the biggest challenge for renewable energy is not generation, but rather how to keep the lights on during periods when the wind doesn’t blow, or the sun shine.
Smart grids are central to solving this problem by managing the energy infrastructure in an intelligent way and so removing the need for polluting power stations to wait expensively on standby.
Artificial intelligence and its cousin, machine learning, enable energy providers to make smarter decisions about scheduling power to different areas of a factory, city or even a country, to ensure that the load is perfectly balanced and everyone can access the power they need.
In fact, utility companies can deploy AI based algorithms to learn how grids behave by analyzing data from a wide variety of sources, including smart meter data, utility-scale SCADA data, EV charging data, and even satellite and street-view imagery. By using these sorts of predictive analytics, companies will be able to identify places where the electric grid is vulnerable to disruption.
Using a vegetation patch as an example, by analyzing satellite and street-view imagery, technology can help to predict the time it will take to grow and also help to anticipate the impact that high winds could potentially have on its growth rate.
That is what DeepMind was discussing with the UK’s National Grid last year, and why Microsoft is investigating how to make power more efficient and predictable. As our energy mix shifts in favour of renewables, and our usage patterns change (not least in the run-up to the various bans on internal combustion engines promised in France, the UK and China), AI-powered smart grids will play a crucial role in delivering this brave new world.
All this is very well, but if AI technology does not have a direct impact on the bottom line it will remain merely a plaything of technologists. The technology industry therefore has to redouble its efforts to communicate on the financial benefits that AI can bring to businesses in the energy sector – and, of course, the consumer.
The fact is that AI has already been proven to pay for itself. When Google bought the British AI and machine learning company for DeepMind for a reported $500m in 2014, the acquisition paid for itself almost immediately. By applying the DeepMind’s “brainpower” to the challenge of its enormous datacenter energy bill, Google was able to shave hundreds of millions off its electricity bill at a stroke.
Given the financial benefits of modern technology applied to the sector, it’s no surprise that investment in North American and European distributed energy companies tripled in value between 2010 and the end of 2016. Meanwhile, investment in big data and AI rose ten-fold in 2017, and accountancy firm BDO found that mergers and acquisitions between energy and AI start-ups rose by a factor of seven – from £500m to $3.5 billion – in the second quarter of 2017 alone.
Artificial intelligence is clearly big business around the world, but it’s not only industry who should be welcoming its arrival. Governments and, most importantly, consumers will gain from greener, more reliable and cheaper energy.
One company estimates that there is six gigawatts of demand-side flexibility in the UK’s infrastructure capacity, which equates to roughly ten percent of peak winter demand and more than the energy produced by the new generation of nuclear power stations that are still years away from coming online. Artificial intelligence is a key technology in discovering how to balance supply and demand, use our scarce resources more smartly, and bring down the cost of electricity at the socket. And, in the end, that’s what every business and consumer cares about most.
By Naveen Kumar, AVP and Practice Head, Energy, Communications, Utilities, Services – Data Analytics Unit, Infosys
