How Artificial Intelligence and Machine Learning Are Changing the Energy Industry

By Chris Buzby

 

Landscape Boiler in stream power plant at night.

Artificial intelligence or AI involves a computer’s ability to adapt to a situation and create a unique solution that was not directly programmed. The energy industry has been tapping into the innovative world of AI to learn how to improve processes, ensure energy resiliency and to enhance the customer experience, among other pertinent goals.

AI vs. Machine Learning
The terms “machine learning” (ML) and AI are often used interchangeably, but ML is just one branch of AI. ML can be defined as a collection of computer science-based methods to find insights by leveraging past experiences to inform a future decision. Over time, these decisions improve and become more precise as the ML system learns from its mistakes. ML itself can be broken down into additional categories and has many different applications.

AI Energy Innovations Currently in the Market
There are many solutions entering the market now based on some level of AI or are enhanced by AI. The simplest form has been around for some time: the Nest thermostat. It learns your schedule through its occupancy sensor, the weather outside, the temperature of the space, your phone or car’s location and the feedback of the individuals changing the set points. After about a week, it knows when to start cooling or heating your home. On top of the thermostat itself, you can also use a smart home assistant to ask the thermostat to change the set point at any time. The complexity of understanding speech and having a seamless conversation with a voice assistant is heavily based on the capabilities that AI has brought to the world.

Beyond the home, we are seeing AI enter every part of the energy value chain. Examples include the work being done at Exelon, such as the generation facilities leveraging AI to better predict failure and optimize when to schedule maintenance. On the distribution side, utilities are using AI-based technology to better respond to storm-related outages. There are many other use cases being explored, and some have already been put into practice.

For commercial energy managers, new solutions that breathe new life into some staple capabilities, such as utility bill review and auditing, and energy forecasting, are becoming available. Additionally, we are seeing some tools complement the role of the energy manager. No longer is the energy manager logging into one of a multitude of dashboards and searching for anomalies; now, they are being notified of an anomaly or an opportunity to create value for their organization.

AI’s Potential to Improve the Energy Industry
AI has the potential to decrease energy customers’ costs and to optimize the many tasks they perform on a regular basis. Cost is more than just price per kilowatt-hour, and energy customers should think about more than just that number when considering energy costs. They should be interested in ways to be more efficient or to decrease capacity costs and other energy-related costs.

For example, through AI, customers will be able to forecast their load as well as cost to help determine the value of energy efficiency and distributed energy (e.g., solar) resources. AI can provide insights into why budget forecasts were off and how that will have a ripple effect through the remainder of the year. In sustainability, AI can help you balance the risk tolerance of your organization with your greenhouse gas-related goals to assist energy suppliers in the creation of a customized commodity product for your business, including offsite renewables and managed commodity products, that meet your needs.

___________________________________________________________________________

About the Author:
Chris Buzby is Senior Manager of Corporate Strategy, Innovation and Sustainability at Exelon, supporting Constellation. He is responsible for supporting the corporation’s culture of innovation, cultivating entrepreneurship and creating internal efficiency through emerging technology. He is currently leading several efforts in the domains of machine learning and blockchain as well as consumer focused applications centered on sustainability. Chris is a fellow at the IDEO CoLab and has lead several cross-industry efforts to explore and commercialize transformative technology.