Energy and utilities providers face a multitude of complex challenges ranging from finding and extracting scarce fossil fuels, to matching volatile supply and demand patterns, to reducing inefficiencies in storage and transmission, to incorporating diverse energy sources into ageing power grids. Although still in its early stages of adoption, AI technologies are enabling companies to meet some of these challenges and optimize the way energy is found, stored, distributed, and consumed.
As computing power, data collection and storage capabilities scale at an exponential rate, AI is poised to revolutionize the energy industry, allowing energy and utilities companies to provide clean, affordable, and reliable power to the communities that depend on it.
The following are just a few areas AI is being used:
Fossil fuel discovery and extraction
On the frontlines, AI technologies such as data analytics and machine learning are helping energy companies discover fuel stores. Oil and gas giants BP and Total, for example, are looking into how AI can be used to uncover insights from their wealth of exploration and seismic analysis data to improve their discovery and extraction processes.
EXXONMOBILE AND MIT COLLABORATE ON DEEP SEA EXPLORATIOIN
ExxonMobile scientists have teamed up with MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) to create self-learning, submersible robots for ocean exploration.
Using AI algorithms, datasets such as flow rates, pressure, and equipment vibrations can be combined with environmental data such as seismic activity and wave heights, to find patterns that help inform, and speed up the decisions being made about anything from the most efficient way to drill wells, to improving equipment reliability, to predicting maintenance requirements.
Electricity generation and distribution
Energy generating assets such as gas and wind turbines and solar panels have always been equipped with sensors, providing real time data for tracking their performance. Today, AI helps companies analyze that data in realtime and take timely actions to prevent outages or risks to outputs.
ABB and IBM partner in industrial AI solutions
ABB and IBM are collaborating to apply Watson’s cognitive insights to the transmission and distribution of energy.
Electricity distribution giant, AES, for example describes AI as an essential component to boosting the visibility, performance and maintenance of its grid systems and assets such as solar farms and gas plants. Advanced neural network design, natural language processing and machine intelligence, meanwhile, are enabling machines to learn on their own, and rapidly analyze data to optimize operations by anticipating and ramping up for surges in demand, dispatching maintenance or technicians to perform precautionary cleaning or repairs in advance, before issues occur and more.
Distributed supply and storage
Energy is traditionally generated and stored in a centralized plant or location, and transmitted to where it is needed, when it is needed, through a power grid. Energy might be stored chemically in lithium-ion batteries, or simply used to pump water up a hill or to compress air in a cavern, and converted back into electricity via turbines when needed. The process of storing and transmitting energy can be both costly and inefficient especially if the location does not naturally have any suitable land formations such as hills and mountains, or if electricity needs to be transmitted across large distances.
Today, with renewable sources such as rooftop solar panels becoming more widespread, and the cost of lithium-ion batteries falling, consumers are now also able to supply and store their own energy, reducing dependency on the grid. This represents both a risk and opportunity for traditional utilities providers. AI helps companies to manage the complex task of incorporating and managing these new distributed sources of supply and storage, and helps define new payment models for customers who perhaps now generate most of their own power needs but still want the security of being able to tap into the grid if, and when, needed.
Renewables such as wind, solar and hydroelectric power will reduce the pressure on scarce fossil fuels, however utility providers must still contend with the challenge of matching volatile supply and demand.
By quickly analyzing data from power generators equipped with smart sensors, AI enables operators to improve resource allocation, while deep learning may allow applications to machines learn independently. By using AI to monitor smart meters and synchrophasers, operators will be able to constantly monitor electricity demand and the flow of electricity through the grid in real time, and actively manage and avoid disruptions. These sensors would communicate with the grid and modify electricity use during off-peak times, thus relieving the grid’s load and lowering prices for customers. For example, on sunny or windy days, when renewables are producing a lot of electricity, AI calibrate the smart grids to reduce the usage of fossil fuels.
Just as with chatbots in the non-commercial sectors, AI is not only helping to answer customers’ common technical and non-technical questions, cognitive computing is enabling it to train itself to improve its responses the more it is used. This frees up the service team’s time to focus on more complex issues and provide greater value to customers.
The energy industry are entering a dynamic era of change and uncertainty. While analytics has long been used in the industry, companies often continue to operate around legacy technologies which lack the flexibility to incorporate newer AI techniques such as machine and deep learning to optimize performance. Although still in its early stages of adoption, AI technologies are enabling companies to optimize the way energy is found, stored, distributed and consumed. To stay relevant in the modern energy landscape, companies will need to embrace AI technologies to be nimble enough to intelligently adapt their services, incorporate new alternative energy sources, and manage their assets and grids cost-effectively.