As the Energy Industry Transforms, AI Automation is Its Next Opportunity 

The global energy system is in transition – from renewable energy generation to electric vehicles – and this is leading to an increasingly diverse, integrated and electrified power network.

But it comes at a cost – to reach that goal of greater global energy sustainability, some $1.3 trillion in clean power generation investments are necessary through 2050, according to the International Energy Agency, which works with governments and industry on such issues.

Fortunately, still-evolving artificial intelligence technologies can be critical tools for this transition by helping to integrate flexible demand and achieve greater efficiency and reliability at lower costs.

But while AI has the potential to accelerate these goals, there are significant barriers that must be addressed to allow the transition to be successful.

As more renewable energy sources are added to the power grid, balancing and managing this increasingly complex system is a huge challenge. One forecast predicts an increase in power system costs by six to 13 percent in 2040. AI can help better balance and orchestrate the network by using AI-enabled decisions in optimizing system build-out, while improving battery optimization and energy management. This would allow a smoother transformation to a carbon-neutral system while better managing costs. However, without an investment of time and resources into AI, this balancing act will not be possible.

Today, we are working with ever-smaller energy domains with increasingly localized control of energy balancing. To move to regional controls of the system, much higher levels of automation are needed. For example, distributed energy resources can be aggregated into one power unit structured as a microgrid or virtual power plant. This “system” will then need to interoperate with other local “systems,” greatly increasing the diversity and complexity needed to manage and balance the grid. AI provides the opportunity to evolve towards a fully integrated “system of systems” in the future, with automation reducing this complexity.

The greater the complexity to be managed, the greater the value of automation. By driving the adoption of renewable resources, we are actively trying to reduce carbon dioxide emissions from gasoline and diesel engines, industries, buildings, homes and more to dramatically reduce air pollution and aim for decarbonization as a goal.

Renewable resources are intermittent by nature and hard to predict and can easily reduce inertia and destabilize the system. To react to this, we need highly accurate weather forecasts and the ability to predict power production levels. At the same time, we need to predict and manage the demand side of the power system by controlling loads such as buildings. For these requirements, AI can help orchestrate the balancing of supply and demand.

In the grid of tomorrow, AI can help us decide which actions to take – like helping us to predict storm-related damage and project where we can generate energy to recover from that damage.

As another example, large logistics companies like FedEx already use AI to optimize the routing of vehicles in their fleet. If FedEx knows how far a driver is traveling, they can optimize the charging of EVs accordingly and migrate more easily to an electric fleet. This provides the opportunity to charge just enough for the day’s deliveries, or overcharge and sell the extra energy back to the grid at the appropriate time. Leveraging AI in this process helps to hedge purchase decisions, optimize operations and even generate revenue.

AI acceleration and adoption is urgently needed to achieve carbon-zero energy use and avert the long-term impacts of climate change. As our energy system increases in complexity, AI will be required to enable us to keep assets stable and reliable, and for electricity to remain a right rather than a privilege.

Digitalization, Automation and Economics

To achieve deep decarbonization, it will be necessary to shift swiftly to an energy system with close to no carbon dioxide emissions. Digitalization is the enabler to this, allowing for the automation of complex processes – like properly predicting the maintenance of digital assets ahead of time and facilitating information sharing within the energy sector.

Like an autonomous vehicle’s AI driver assistance system can recognize a road from another car and applies logic to determine the next best step, AI can rapidly sift through troves of data within the energy sector to identify patterns, calculating how to best respond to anomalies and initiate proper action. Removing the unknowns that previously were not analyzed and acted upon as efficiently quickens the pace of the energy transition, which is economically beneficial for the levels of investment required.

According to the BNEF’s New Energy Outlook 2020, 56 percent of power generation could be provided by solar and wind in 2050, assuming no further policy support from today’s levels. This would require investments of $5.1 trillion in solar, wind and batteries and a $14 trillion power grid investment by 2050. Hosting more renewable power increases complexity, and operations can be improved through streamlined processes, enabled by automation. The realities of 2050 require those investments today.

Consider the lifetime of power grid equipment. Without intervention, increased air temperatures due to climate change could reduce the lifetime of power grid equipment and transformers by 10 years, equating to an additional $188 billion in replacement costs. AI can help operators avoid this additional cost by keeping transformers within optimal operating ranges, but this requires digitalization.

Even if AI were to reduce the cost of investment or energy demand by a small percentage, this would still drive billions of dollars in savings for the industry and for consumers.

AI Has Many Roles to Play

Harnessing AI to accelerate the energy transition – through equipment, sensor data, images, videos and market, commodity and weather data – is critical. Automation provides the most value where there is the highest level of complexity, from network control systems to generation plants to the strategic planning of corporations.

The energy industry would benefit from approaching AI-related technology governance in a proactive and collaborative way, and the coming years will be crucial in unlocking the opportunity. By adopting common data standards and implementing digitalization more broadly, it becomes possible to capture the full spectrum of AI opportunities.

About the Author

David Goddard is the head of digitalization for Hitachi Energy and is responsible for driving the company’s digital evolution. He formerly worked for Cisco Systems in various executive leadership roles, including being responsible for the company’s global IoT practice and for running the Cisco Customer Assurance crisis response organization. Goddard earned a bachelor’s degree in telecommunications and electrical engineering from Farnborough College of Technology.