Stephen J. Callahan, Vice President, Energy, Environment & Utilities, Global Strategy IBM Industry Academy Member
It is a huge understatement that leveraging technology—analytics specifically—is the foundation for advanced asset management in asset intensive industries and in the utility industry in particular.
The business outcome is clear. McKinsey estimates that leveraging advanced analytics techniques in the utility industry can reduce the cost of maintenance by 10 to 40 percent across multiple asset families. Beyond maintenance cost reductions customer satisfaction improvements, better regulatory compliance, and increased reliability are also frequently realized benefits.
These advanced asset management practices are enabled by the cornucopia of data rich technologies that are fast maturing. Ranging from industry specific examples such as cheap and attribute deep line sensors, advanced meter (AMI), and advanced geo locational information such as GIS, LIDAR and satellite imaging to the explosion of IOT devices spawning rich and useful video, audio and conditional data, leveraging this plethora of data is at the core of these asset management practices.
One example is applying analytics against AMI data leads to significant improvement in operational awareness. Prevailing methods of determining which customer is powered by which phase of which distribution transformer, as well as determining interconnections between laterals and transformers can be time-consuming and costly when conducted in the field. Utilities are reliably identifying problem areas with 90 percent accuracy without the need to send crews into the field. This can be achieved by applying analytics to smart meter data to determine which meters are connected to which electric phase, then providing recommendations to fix connectivity records.
Or consider the strategic requirement to accurately forecast renewable output. To be economically competitive, wind farms need to maximize the availability, reliability, and performance of their assets. Analytics provides situational awareness of wind farm assets using historical and real-time data, including weather. This allows wind farms to monitor wind speed, blade angles, vibration, yaw speed, power AT, and wind direction, accurately forecasting output, while providing advanced warning of pending asset failure.
“Outcomes are now achievable that lead to operational changes based on prediction. Asset management is no longer necessarily reactive”
These are not hypothetical examples. They are operational today. These and many other examples illustrate a state of the art of analytics implementations that leverage advanced data science methods coupled with data that is informationally deeper and more real time enabling outcomes that are more operationally excellent and more competitive in the marketplace.
The strategic factor in all these examples is data. High quality (accurate, consistent, and curated) data enables utility enterprises to substantially increase the sophistication and efficacy of asset management. This has been the path of the industry’s evolution as enterprise asset management (EAM) is increasingly supplemented by asset performance management (APM) which is intrinsically predictive in nature.
Ginni Rometty, IBM’s CEO, has called Big Data the “World's natural resource for the next century.” The Economist recently echoed this insight by heralding data as the new oil.
From a business perspective these advances have enabled utilities to elevate operations to conditional based management. This is no insignificant accomplishment. It is an advance from the historical temporal (periodic maintenance) or episodic (car hits pole) mode of asset management. Outcomes are now achievable that lead to operational changes based on prediction. Asset management is no longer necessarily reactive (See Figure 1).
But this is not the end of the story. It is the beginning. A new era is afoot.
AI is mainstreaming, and it is going to rapidly enable “pre-conditional” asset management. Consider the potential. Traversing the vast store of conditional data now being routinely collected by utilities, advanced AI models can scan the data to find antecedent events—ones that do not themselves create catastrophic failures, but that—collectively, sequentially, and over time create such failures. Avionics, among other asset intensive industries, have already realized such capability in jet engine inflight management and utilities are not far behind.
IDC predicts that by 2020, 25 percent of utilities will integrate asset performance management investments with sensor data and cognitive capabilities, boosting asset efficiency and reducing maintenance costs.
Leaders are leveraging cognitive technologies are producing favorable business outcomes. An IBM global survey of more than 6000 organizations finds that organizations are using deep analytics: 74 percent are pervasively using at least one type of prescriptive analysis. A small group of the organizations we surveyed—representing 4 percent of the global marketplace–are blazing a trail on their cognitive journey. We call them the Cognitive Explorers. They are found in each of the 19 industry groups we surveyed across all geographic regions. And according to respondents, these organizations are all-around outperformers: 84 percent report they outperform competitors in revenue creation; 91 percent report they are more effective at what they do than competitors; 89 percent report they are more profitable than similar organizations; and 89 percent report they are more innovative than most in their industry.
The bottom line is cognitive analytics can help your organization gain a competitive advantage especially in venues where the competitive threat of industry disruptors makes managing assets not just good practice, but instead critical practice to thrive in the new energy ecosystem.
So why love your data? Because it is the key to animating the power of the AI based analytics that are soon to be the core of your operations. The utility industry is historically, some would say legendarily, poor at leveraging its data. The historical utility perspective views data as a burden—something to be stored at a cost—due to regulatory requirements. Monetizing it faces hurdles seemingly insurmountable.
But such pessimistic views miss the potential. Forward-looking utilities know that data, leveraged in asset management and many other internally-focused functions, IS your competitive advantage. If you doubt that consider how many new entrant disruptors are after it.