What is the 21st-century electrical grid? That question is certainly getting a great deal of attention these days. Conferences, seminars, and associations are providing a constant source for speculations and predictions as to what the next generation of the grid is going to look like. This grid is changing to have better operational performance, offering more options, opened to more competition, providing access to renewable resources, and a changing marketing structure. Today’s grid is enhanced with embedded digital technology in the form of sensors, software, and monitoring systems.
All are interconnected by high-speed communications networks, which integrate real-world objects into computer-based systems, producing real-time data. It’s called connectivity. And it is every bit comparable to the Internet of Things (IoT). Some experts have gone so far as to call this the Utility Internet of Things (UIoT) to bring commonality to understand what is going on. Digital technologies are reshaping the everyday characteristics of the grid into this UIoT, but taking advantage of the technology brings another element that defies common understanding: big data! It’s not an exaggeration to say big data is enormous. It is data that is too large or complex to deal with by traditional data processing methods.
Rise in AI Interest
To give an idea of the amount of data being discussed, a recent Forbes article reported, “In the past two years, we created nine times more data than we did between the dawn of man and the year 2015.” That is a mindboggling statement and implies that it is beyond human capacities to deal with data that large, which is true. To paraphrase companies like IBM, Microsoft and Oracle, data this large has to be processed and leveraged to provide intelligent insight to be useful.
Specifically, the data owner must be able to manipulate that data, but that is proving to be a challenge. Utility Analytics Institute (UAI, the sister brand of T&D World) conducted a survey of 42 major utilities focused on an assessment of the analytics maturity taking place within the utility. UAI asked the utilities to rate their company’s ability to process big data with the tools and technologies available to them. After reviewing the results, UAI said, “In terms of whether analytics toolsets meet enterprise needs, most responses fell in the poor/limited and good categories.” Clearly, there is a need for user-friendly methods for managing big data.
Managing big data isn’t impossible, and there is a technological solution getting attention known as artificial intelligence (AI). AI has the ability of taking big data, boiling it down and making sense out of it. This fast-paced technology has moved to center, but adoption has been gradual until recently. In early 2019, Gartner Inc., a research and advisory company, published its 2019 CIO Survey. More than 3000 chief information officers (CIO) from 89 countries across major industries took part. The Gartner report said, “Four years ago, AI implementation was rare, only 10% of survey respondents reported that their enterprises had deployed AI or would do so shortly. For 2019, that number has leapt to 37% — a 270% increase in four years.”
To put this in a more tangible perspective, the International Data Corp. (IDC) published its Worldwide Semiannual Cognitive Artificial Intelligence Systems Spending Guide last year. IDC said, “Worldwide spending on cognitive and artificial intelligence systems will reach US$19.1 billion in 2018, an increase of 54.2% over the amount spent in 2017.” The report went on to forecast, “cognitive and AI spending will grow to US$52.2 billion by 2021.”
In spite of having trouble understanding AI, we are using it. However we may not be totally aware of how pervasive AI has become in our lives. Google Assistant, Alexa, Siri and several others are intelligent assistants fighting for our attention. Even our cars have voices that can be actuated to let loose a host of functions like GPS navigation systems. So, in one sense AI is being used by a lot of people for a lot of functions. Since AI is here to stay, it is important to have a basic understanding of what AI can do and what its limitations are.
What is AI
One of the biggest confusion factors is all the different terms that are used as synonyms for AI such as machine learning, deep learning, cognitive computing, etc. The list grows daily. Keep in mind, these terms are not interchangeable, but they are often used that way. That doesn’t help anyone trying to figure out AI or how to use it. First of all, AI is a division of computer science using complex instruction sets to perform what appears to be human-like intelligence. These programs are powered by algorithms, and that is the ingredient causing the mystique. Without going into a lot of detail, an algorithm is a set of step-by-step computer instructions that can use data to build models that make predictions based on the data. Remember, we are a long way off from the thinking, talking robots seen in movies and on television.
Algorithms are how AI demonstrates being smart, but be aware it’s not intelligent, which is the critical distinction. This type of AI is referred to as Narrow AI or Applied AI. It is said to simulate human thought, but each application can only carry out one specific task with a limited range of functions. An example of Narrow AI is found in many applications used today. Each application is capable, but only for a specific task. Case in point, a chess program plays chess, but can’t manage assets. An asset managing application doesn’t play chess. Narrow AI applications do, however, take data, manipulate that data, find patterns, and turn the data into useful information that can anticipate needed action.
Fact or Fiction
If you are interested in AI technology that simulates human thought and can carry out multiple tasks, then the Broad AI or Generalized AI type meets these requirements. The Broad AI application plays chess, performs facial recognitions, drives a car and performs a slew of other tasks. It is capable of multitasking. Broad AI performs on a level equal with a human, but there is one problem: It is not available yet.
Broad AI is in the fictional category right now, but it is the direction developers are going. Since the fictional category has been mentioned, there is another type of AI that usually pops up whenever AI is discussed. It is known as the Super Artificial Intelligence, which right now is science fiction. It’s the AI slated to surpass human intelligence in all aspects, but before worrying about it, the Broad AI technology needs to be developed, deployed and all the bugs worked out.
As stated, the current capabilities of the AI technology are limited to the Narrow AI variety, which is more about data mining and producing valuable information. Fortunately, this is exactly the AI subset needed for the modern electric power grid. Last year, Navigant Research reported, “More advanced UIoT analytics solutions have entered the market and can be applied to legacy systems and new data flows using edge computing, cloud computing, machine learning, and AI to unlock valuable insights and drive operational efficiencies.” These AI-powered platforms are getting attention and manufacturers are bringing more of them to market.
The Tool Box
A recent release from Research and Markets, a marketing research company said, “Artificial Intelligence is one of the biggest technological trends transforming every business sector across the world.” It went on to say that AI possesses tremendous potential to transform the energy and utilities sector. In combination with other technologies like Big Data and IoT, it can aid the active management of electricity grids by balancing demand and supply.
Interest is high, and research is taking place in a variety of AI-powered applications for the grid. Some of the most promising are in asset management, demand response, outage management, customer services, energy storage, renewable resources, and many other areas in the power delivery system. Probably one of the most successful applications using AI technology is integrated asset management systems (AMS). This is an extremely important area for utilities, since sharing asset information throughout the organization has been identified as a method for improving enterprise efficiencies.
AMS also have provided some significant impacts in operating and maintenance cost reductions for the utilities taking advantage of them. These systems are designed to provide utilities with end-to-end asset tracking, which includes predictive maintenance programs, failure analysis platforms that reduce risks, and asset health awareness systems to increase reliability. Manufacturers are providing AMS such as ABB’s Asset Performance Management: Ellipse, Bentley’s Asset Performance Software, GE’s Asset Performance Management Powered by Predix, IBM’s Maximo Enterprise Assets Management platform, Oracle’s Utilities Work and Asset Management and Siemens Reliability Centered Asset Management.
Distributed energy resource systems and renewable energy resources are getting a boost from AI too. Early last year, the Department of Energy announced new research efforts by their SLAC National Accelerator Laboratory at Stanford University using AI applications to help utilities better integrate solar resources into the grid. SLAC came up with a “first of its kind software platform” called VADER (Visualization and Analytics of Distribution Systems with Deep Penetration of Distributed Energy Resources). DOE said, “The VADER platform can model potential changes in connectivity and the behavior of DERs on the grid, enabling the real-time optimization and automation of distribution planning and operation decisions.” This is a handy application when it comes to photovoltaics and the power fluctuations they cause.
Several companies are working on combining AI technology and energy storage. Stem Inc. installed two AI-powered energy storage systems at two California malls totaling 4.4 MWh. Con Edison along with Peak Power Inc. (Synergy intelligent software), and Lockheed Martin (Gridstar 2.0 energy storage technology) took part in an energy storage system using AI technology in White Plains, New York. The system is a 375 kW/940 kWh battery energy storage system. The AI system optimizes the operation of the battery storage for a commercial building to reduce energy costs for the system.
No matter how you look at it, AI will change the grid as more AI-oriented technology is integrated into the power delivery system. It will become a fundamental part of the UIoT where smart grid components communicate with each other and all parts of the enterprise.
It’s an exciting time and a scary one too. Remember when the industry had to develop and adopt rules and standards for interoperability? Many experts are cautioning that the industry is in that same place when it comes to AI. It’s going to be a challenge, but that is what makes it exciting and why understanding the abilities and limitations of AI technology so important.
Editor’s Note: This is the first article in a series focused on how AI is being applied on the power delivery system. Future articles will explore the applications in greater detail.
