Operational Technology innovation embodied in the Industrial Internet of Things (IIoT) is proliferating at a rapid pace. To date, the electric power industry has not taken full advantage of this sea change. Accenture predicts that IIoT will add $14.2 trillion to the global economy by 2030 (https://www.accenture.com/us-en/insight-industrial-internet-of-things), while experts like Steve Bolze, president and CEO of GE Power, claim that the world will need 50% more power in roughly the same time frame (http://news.mit.edu/2016/ge-joins-mit-energy-initiative-develop-advanced-technology-solutions-transforming-global-energy-0830).
Renewable assets and distributed energy resources are playing an important role for utilities as they seek to meet increased demand; however, leveraging IIoT data and analytics is an equally important, critical component of grid modernization. There are three relatively simple ways to improve upon existing infrastructure and enable use of IIoT technology: first, introduce a Common Information Mode (CIM), secondly, employ a real-time dataflow engine, and finally adopt cloud computing infrastructure.
At present, OT architectures do not provide contextualized event management or actionable intelligence. Typically, OT architectures allow devices to publish only a device ID and device data to a utility’s data network. No information about the role of the device or application is included with basic telemetry data. Centralized management via GIS data or a Distribution Management System (DMS) produces limited data. Utilities working to make the most of IIoT technologies must be able to provide data richer than device ID and basic device measurements.
To create more robust data and enable distributed computing, utilities should employ a Common Information Model (CIM). CIMs establish a richer data model that codifies the relationship between data and helps to manage diverse models and naming schemes across systems and devices. CIMs define the semantics of data and use these definitions to establish a unified web of information without the overlaps and conflicts traditionally found in Operational Technology integrations. Exposing consumable data and performing the necessary classifications so that contextualized data is available to the appropriate smart grid application is important; however, utilizing a CIM alone does not create the necessary real-world connection to the virtual world.
When working to establish a real-time analytics program, an existing ‘real-world’ physical connection, frequently referred to as a ‘Y gateway,’ must be exploited to create an exposed model. The most efficient way to establish and exploit ‘Y gateways’ is through the use of architecture designed to do so. LiveData Utilities refers to this architectural pattern as Operational Technology Message Bus (OTMB). OTMB expands upon traditional architectural patterns employed by utilities. OTMB acts as a real-time utility protocol-aware dataflow engine – providing SCADA-class in-memory processing, configuration and manipulation of dataflows at run time, and seamless integrations to IT and OT systems. OTMB architecture pattern is capable of handling diverse datasets from a nearly unlimited number of data sources, homogenizing the various outputs of a utility’s stack, which, in turn, allows the data being created to be consumed by a wider variety of systems and applications. The use of these Y gateways established using OTMB architecture allows utilities to share data sets to open source tools like historians or machine learning and analytics engines; this is a critical capability with the growing popularity and rapidly improving performance of cloud computing platforms and infrastructure.
The question of how to implement and leverage IIoT remains. Today, utility asset analysis looks at a small number of cross correlations. The reality is that this approach is outdated and ultimately does not take advantage of existing data already captured by utilities. Utilities need to be smarter about applying cross correlation and other machine-learning techniques across a much wider breadth of grid systems. While utilities are already suffering from a data deluge, data volumes are continuing to grow as new data is stacked upon old data. While a CIM model exposes usable data and elevates data beyond simple timestamp recordings, and Operational Technology Message Bus architecture exploits gateways back to the data network; data in itself is not a resource that creates meaningful, actionable insight – and neither more data nor ‘Y gateways’ help to manage the data tsunami.
Public or private cloud-enabled computing is a key characteristic of IIoT. By adopting cloud computing infrastructure, rich data analysis can be performed on Operational Technology data – delivering actionable insights without the need for an in-house data scientist. The richest source of actionable insight is via these cloud computing platforms, which have been developed to recognize deeply entrenched patterns that would not be detected by human analysis. Operational technology data with a Common Information Model applied can be queried with the depth and specificity needed to derive actionable intelligence from pattern recognition and other machine-based data analysis.
Ultimately, this enables operators to achieve a deeper understanding of exactly what is happening within the grid, where it is happening, and most importantly, why it is happening. In the past, utilities have been reluctant to employ cloud-enabled solutions because of utilities’ unique real-time requirements. IIoT is designed to address real-time requirements, but to take full advantage of IIoT technology, utilities must support architectures that apply cloud-enabled solutions.
We work with utilities across the globe - every utility I work with has a unique collage of devices, applications, databases, Operational Technology and IT resources. The recent proliferation of grid edge technologies has resulted in each piece of the technology collage creating data in its own way – without exposure to other parts of the collage or the data created within those applications and devices. Only centralized computing is possible as a result of this diversity found in smart grid data. Common Information Models not only homogenize application data, but also elevate that data from being basic historian data to being the foundation for operational analytics.
Operational Technology Message Bus architecture and similar OT dataflow engines allow for necessary ‘Y gateway’ connection to devices, systems, and applications, as well as provisions for OT data hundreds of times per second. Additional benefits of employing OTMB or similar architecture include minimizing total configuration efforts and allowing for partitioning of systems, which is important when managing OT/IT hybrid scenarios found in many smart grids.
As more utilities integrate distributed energy resources and grid edge technologies, a performant architecture, a Common Information Model, and the introduction of cloud-enabled solutions are three additive updates to utility grids that will jumpstart IIoT deployments and ultimately help transmission, distribution, and generation suppliers meet growing demand and optimize current operations. The capability to deeply understand activity within transmission and distribution grids is the very essence of why IIoT is a critical component to modernizing our aging grids.
About the Author
Adam Reiss is Senior Marketing Associate at LiveData Utilities, managing the marketing activities for utility products and services. With a background in institutional research that covers everything from investor statements to coral reefs, Adam has a passion for data, and enjoys delivering actionable insights from unlikely sources.
