Part I: The Technology and Data
By Paul McRoberts, President
I took my car in to have the oil changed the other day, and, naturally, it reminded me of how far the utilities space has come, especially on the analytics side. Maybe I’m dating myself here, but I come from a time when we changed the oil in our cars after a set number of miles, no questions asked. Why did we do it that way? Because, based on manufacturer testing, that was the finger-in-the-air estimate of when the oil needed to be changed.
My most recent experience was different, however. I didn’t take my car in because it had reached an arbitrary number of miles—I took it in because the car told me the oil needed to be changed based on how it was performing. We’ve reached a point where technology can be deployed to collect and respond to available data more intelligently than ever before, giving us the insight to make better decisions and investments. Clearly this is the case in automobile maintenance, and it’s certainly the case in utilities.
In part one of our “What's Disrupting Utility Analytics” series, we explore the evolution of the utilities space, starting with disruptions to technology and data and how they’re each being leveraged to run infrastructures.
Let’s take a closer look at the utility analytics disruptors on the technology side.
- Smart sensors and continuous data collection
More specifically, I’m talking about smart sensors that are ubiquitously deployed anywhere across distributed infrastructures—all thanks to the internet of things (IoT) explosion.
The utilities space has arguably been the greatest benefactor of IoT technology. I’ve even seen the EIA estimate that utility companies will have over 1.5 billion connected devices under their thumbs by the year 2020. That’s a lot of smart devices. More importantly, that’s more data being continuously collected than ever before in history, leaving stakeholders and plant managers to answer the inevitable question of how to best utilize it.
- Advanced pattern recognition
With record amounts of data pumping in by the second, we should be able to mine information that we never had insight into before, right? On paper, yes. But we’re talking about billions of bytes of data being collected in real time from a variety of different sources—even a team of engineers working 24 hours a day for 365 days straight wouldn’t be able to analyze and respond to this amount of data.
Fortunately, technology has kept pace with sky-rocketing volumes of data in the form of machine learning and artificial intelligence tools that can aid us in recognizing advanced patterns rippling through our networks that had previously been invisible. Maybe fuel consumption spikes across a certain asset class at a specific time of year, or maybe a compressor is operating outside the boundaries of normal, which is then causing anomalies in a generator downstream. Having the ability to analyze these patterns as they’re happening allows us to take a massive leap forward in not only responding to issues across our networks but predicting them.
- Risk management
In the utilities space, we prefer to think of risk management as an opportunity, not a danger area. And with advanced pattern recognition capabilities, risk management shifts undoubtedly in favor of being an opportunity. Noticing an issue with an asset and being able to anticipate how it will impact other areas of our systems down the road changes the conversation around risk management. Not only are we able to assess and respond to potential risks faster and more holistically, but instead of risk management being an effort to respond to future alarms (“What are the odds something bad will happen in my system?”), it becomes a strategy for getting ahead of them (“What are the potential costs of not doing something to improve my system?”).
From Reacting to Predicting
This evolution in risk management is indicative of the more universal disruption that’s happening across the utilities space, and that’s a shift from reactionary operations to predictive ones. This idea should excite stakeholders everywhere because it translates to getting better returns on our assets. Now we can look at an isolated issue happening with a generator, predict how that issue will affect other assets downstream and take steps to mitigate that affect, optimize performance and safeguard our investments. Instead of resolving an unplanned shutdown, we’re now utilizing our planned shutdowns in more beneficial and innovative ways.
Check out part two of our series, where we take a closer look at how evolution in technology and analytics is leading to further disruptions within our staffs and skillsets.