In spring of 2018, JEA reemphasized its efforts to determine the best path forward towards a condition-based asset maintenance program. What emerged from a myriad of viable options was a scalable, sustainable IIoT (Industrial Internet of Things) Platform, powered by intelligent edge devices and robust data handling capabilities.
Like all utilities, JEA is evolving into new business models as consumer demands and the technology landscape changes at a staggering pace. One of the evolving business models of critical importance is asset management, and maintaining the multi-billion dollar portfolio of assets supporting electric, water, and wastewater service operations at the 8th largest municipal utility in the US. With a vast service area of over 850 square miles, JEA turns to technology – some legacy, and some emerging – to add efficiency, awareness, and transparency to our asset management processes.
The development of JEA’s IIoT Platform has been long in the making. In the late 1990s, JEA implemented a centralized enterprise time-series data historian that integrated with various plant SCADA systems to capture and share asset measurement data. This data historian evolved and grew over time, becoming critical to JEA Operations. Data collection was buoyed by the ownership of our own communications infrastructure: our ringed fiber architecture helped to enabled reliable fast Ethernet connectivity, and allowed distributed data collection from various sites into a centralized server. Having this in place enabled JEA to do process control optimization using a combination of AI/NN and ML models – use cases included boiler optimization, well field pump optimization, and water operations optimization – but this was only the beginning.
In conjunction with our technology focused Enterprise Information Management program, company asset management leaders worked to identify those use cases that provided exceptional value to JEA, and that would facilitate a shift from reactive maintenance and equipment failures to more proactive maintenance based asset condition monitoring. Among the factors making this shift possible, the market has provided numerous end point devices for monitoring asset telemetry and performance. Moreover, many of these new devices offer compute capabilities, are inexpensive, and ‘plug in’ to established networks via standard protocols already in use. This allows for an exponential increase in the number deployed sensors, the number of data streams from the sensors, and the volume of data collected. In order to take on a change of this magnitude, we embarked on the following course:
1) Identify specific use cases for automated IIoT monitoring, based upon business value and complexity;
2) Deploy IIoT solutions based upon a limited array of technologies to support these use cases;
3) Continue to augment the IIoT platform with new end points and data streams on a use case-by-use case basis – grow the platform incrementally with business-prioritized use cases.
With this simple framework JEA moved into the IIoT space, focused on priorities that drive business value.
Despite an effective approach and a common objective across both business and technology, opportunities remain for JEA’s fledgling IIoT program. Across the landscape of potential sensors supporting electric, water, and wastewater operations, proprietary packaged solutions provided by vendors offer a trade-off. While they are perhaps easier to install and implement, they include their own communications technology (usually cellular), their own hosted SaaS application, and their own data repository. A proliferation of these proprietary solutions leaves the utility with too much to manage – too many networks, too many SaaS portals (some lacking accessible APIs), and too many disparate data stores. All of these factors make centralized data and operational alarm management much more difficult. User expectations are another opportunity for effective management – some business partners expect all data to be readily accessible and to integrate with existing tools, but the reality is that this takes time. One key lesson learned is the focus on data integration and accessibility; these factors empower the users to spend more time doing the analysis and analytics, and less time wrangling with the data from disparate systems.
“In conjunction with our technology-focused Enterprise Information Management program, company asset management leaders worked to identify those use cases that provided exceptional value to JEA, and that would facilitate a shift from reactive maintenance and equipment failures to more proactive maintenance based asset condition monitoring”
Other lessons learned have emerged as well. Our IIoT program guiding principles call for a separation, both physical and logical, of the IIoT system from existing OT control systems. IIoT data collected is indication only, with no control capabilities or outputs. This helps a new program avoid the heavily regulated, reliability-focused network used for established controls. Additionally, legacy OT protocols mapping tag-based licenses are not designed for high throughput, and are unnecessarily complicated for our data collection purposes. Moreover, for the new IIoT network, modern security needs dictate that protocols must support encryption and security certificates. Our IIoT network is focused on data collection, is reliable but not mission-critical, and benefits from reduced regulatory burden, streamlined change management, and more agility when a modification is needed.
Despite these opportunities and lessons learned, the future of IIoT utilization at JEA is extremely bright. Low cost sensors and lowered monitoring costs enable monitoring that was previously prohibitively expensive. Cheap data networks and cloud computing allows collection of data from distributed sources like never before, and intelligent devices continue to spread out at the edge. Our move towards enterprise scale management of the IIoT platform will incorporate new tools (both open source and commercial), opportunities to leverage artificial intelligence and machine learning libraries, and interoperability of systems using open APIs. With almost too many options, the technology is not the hard part! A clear objective shared across the operations and technology has made possible our move to IIoT, and the development of the nascent IIoT platform has enabled new capabilities and improving efficiencies and effectiveness for our utility business
