Originally featured in Intelligent Utility.
By Alyssa Farrell, Global Product Marketing Manager, SAS
The Internet of Things is the concept of everyday objects – from industrial machines to wearable devices – using built-in sensors to gather data and take action on that data across a network. Thierry Godart, General Manager of Energy Solutions at Intel, describes IoT as “taking the best of IT into the operational world.”
IoT devices are indeed flooding utility operations. From synchrophasors to smart meters, new assets installed in the energy grid are highly sensored, due to the declining cost of the sensor technology. And customers aren’t left out either. Home automation technology can sense and respond to the needs of an individual or family, even when they are not present.
Across all industries, Gartner forecasts that 6.4 billion connected devices will be in use worldwide in 2016 – up 30 percent from 2015 – with the number reaching 20.8 billion by 2020. In 2016 alone, 5.5 million new devices will be connected each day.
For utilities, the smart grid era unleashed not only millions of these new IoT devices, but also more data that utilities need to analyze and understand to make better decisions about their networks. In fact, 63 percent of utility respondents in a recent SAS survey indicated that IoT was critical to their companies’ future success.
Top uses today include “metering and meter data management” (55 percent) and “cybersecurity” (49 percent). However, the biggest growth areas are customer-facing. Seventy-three percent of respondents indicated that “customer engagement” will be IoT-enabled in the next three years. It should come as no surprise that the most common benefit of IoT analytics cited in the research was better customer service.
While IoT devices are largely grid-based, utilities are turning that intelligence into benefits for customers. For example, Kevin Lagge, Director of Strategy, Analytics, Enterprise Architecture and Technology Planning at Oklahoma Gas & Electric, has a real passion for applying analytics in the area of reliability and outage management. IoT is influencing the discipline of grid reliability by integrating data from social media and communicating with customers via their mobile devices. “As we get more sophisticated with our customers and we’re leveraging more about what they’re telling us, we are going to be able to serve them better and increase customer satisfaction,” Lagge said.
Our research also probed priorities in the area of machine learning. Only 20 percent of respondents indicated that they have a good sense of what machine learning can do for them.
IoT is ahead of machine learning in terms of familiarity and adoption in utilities, but it is important to consider both as a partnership for delivering a more self-sufficient grid. “I think they’re very, very linked,” said Raiford Smith, former Vice President of Corporate Development and Planning at CPS Energy. “You can have one without the other, but they feed off each other.”
The most common benefit of machine learning was cybersecurity. Across the board, survey respondents were more likely to associate machine learning with grid-related data analysis. What we infer from these divergent opinions is that IoT excels at connecting devices, particularly at the customer level – including better integrating distributed energy resources and improving customer engagement – whereas machine learning shines when it is strategically applied to analyze, adapt and learn from data coming from those connections.
As early as 1959, Arthur Samuel defined the concept of machine learning as the ability of computers to learn to function in ways that they were not specifically programmed to do. The philosophy behind machine learning is to automate the creation of analytical models in order to enable algorithms to learn continuously with the help of available data.
Machine learning can be applied in cases where the desired outcome is known (guided learning), or the data is not known beforehand (unguided learning), or the learning is the result of interaction between a model and the environment (reinforcement learning). This makes it a perfect marriage for data from IoT devices.
“There are going to be so many IoT devices out there that we’ll have to develop machine learning or analytics that can take into account the sheer volume of data,” said Lagge. “It’s critical, from the standpoint of the volume of information we’re going to have in the future, that machine learning is developed. We’re not going to be able to have a human interface every time we want something to improve.”
As utilities worldwide move more into a connected state of business, key partners will become increasingly important to help analyze the wealth of data to make smarter decisions for both the grid and customers.