DRINKING THE BENEFITS OF BIG DATA IN WATER
Big Data. The term conjures up visions of teams of scientists, analysts, researchers, and practitioners poring over vast sets of infinitesimally granular data to derive fabulous correlations that bring insight into our macro world. Certainly the use of “big data” in fields like medicine, genomics, physics, astronomy, finance, sociology, and energy exploration has yielded significant benefits—new drugs, the Higgs boson, exoplanets and genetic patterns. These discoveries would not have been possible without the mathematical and statistical analysis of big data. Despite the same significant potential for the water industry, our utilities have lagged these other industries as it relates to data collection, curation and use.
THE BIG DATA DROUGHT
The reason for this lag in adoption of big data by water utilities is due in large part to the structural confines of our highly fragmented, and low scale industry. While many consider water utilities to be large organizations with significant resources, there are in fact only 244 public water systems in the United States that serve over 50,000 connections.1 Over 99.5 percent of our utilities are truly small organizations—and that does not include the additional 100,000 transient non-community water systems (TNCWS) and non-transient non-community water systems (NTNCWS).2 As such, the scale required to really implement true big data is almost never achieved.
Some water utilities have begun to embrace the promise of data. It has been, however, limited to investment in control systems and Supervisory Control and Data Acquisition (SCADA) systems. Notably, however, these systems are rarely used for projections, analysis or discovery—rather they are traditionally used for reporting and automation. And getting data out of SCADA systems can be a challenge in of itself. Most water utilities do not have the resources to invest in sensors and automated data collection systems, let alone the platforms and skills necessary to analyze this data. With the recent significant financial pressures as well as the pressure to consider the effects of external influences on water supply and delivery, it is time for water utilities to evolve.
Our water utilities need to undergo a fundamental change. Driven by increasing water volatility, increasing financial pressure, increasing customer expectations and increasing infrastructure costs, it is clear that the way water utilities have operated in the past will not be the way they will operate in the future. Our highly regulated industry is being asked to meet increasingly tough regulatory standards while keeping costs low—and these two outcomes are progressively more mutually exclusive.
Water utilities have one of the most direct impact on the health and well-being of the public in the United States, second only to the food industry. And while all of us will look at the “best before” date on a gallon of milk before we drink it, when it comes to water we open the tap, expecting our water to safe, clean and present. We trust that the water utilities are managing the water in our distribution systems such that it is never past its “best before” date.
Now, such a reliance is a well-founded assumption. Here in the United States we are provided with very high quality water—at all taps in our houses, at all times. However, as our infrastructure ages, and our utilities continue to be financially challenged, the potential for our fortunate situation to change for the worse ever increases. In fact, there has been an increase in waterborne diseases attributed to the condition of the distribution system over the past 30 years.
As such, for water utilities this is much more than an academic, operational or financial problem. The reality that utilities face today can have real public health impacts. Our water utilities need the tools to manage their systems in a much more proactive way. That means they need data. More accurately, they need more insight from data.
BIG DATA BENEFITS
Using large data sets and analytics, water utilities can begin to actively manage their operations in response to financial distress and degrading infrastructure by:
- finding and protect revenue;
- finding and fix leaks;
- understanding the real time condition of infrastructure;
- predicting demand;
- reducing operating costs through dynamic pressure control;
- predicting the quality of water in their distribution system; and
- engaging customers in their water use.
And that’s just the beginning.
In fact, with the availability of advanced metering infrastructure (AMI), the rapidly decreasing cost of sensors, and the availability of massive data sets, including weather and climate, hydrological, and remote sensing information, means that we have a significant opportunity to expand the data footprint for utilities. This convergence of technologies is the promise of the Internet of Things (IoT).
In order to realize this promise, however, we need to find a place for that data to land, develop a curation mechanism for that data, and provide the means to analyze that data. In addition, we need a mechanism through which our utilities can enjoy the operational, financial and customer benefits of intersecting and interacting with this data without incurring out-of-proportion costs—it must be easy and economical and most of all, not detract from the primary objective of water utilities to provide safe water to end users.
SLAKING THE THIRST FOR BIG DATA
The reality is that most consumers of big data are not directly interested in the data, but rather they are interested in an outcome or actionable output from that data that can drive a decision or further analysis. For water utilities, finding a means to economically connect distribution system sensors, external data sources, and internal data sets with built-in or configurable analytic tools is one way to maximize the adoption and functionality of big data tools.
The unique data architecture and interface of FATHOM provides water utilities a one-stop shop for data and analytics and provides the means to truly understand the inferences of the data. The FATHOM time series database is structured to allow for the ingestion of data from any meter or sensor, while application programming interfaces (APIs) allow for access to external data sets and analytic tools to derive insight from that data. These services are available at low cost through the FATHOM software-as-a-service delivery mechanism.
In order to create truly actionable outputs, we need to not only consider the data but also the context of the data. FATHOM—born from water utilities—is a leading subject matter expert in this regard. The people working in FATHOM have designed, built and operated water utilities and have extensive experience in the meaning of the data. As such it’s not only the data that drives the analytic output—but our experience. A real-world perspective to correctly interpret the data.
Employing FATHOM, water utilities can maximize the usefulness of their own data, access external data sets and use built-in analytics to derive value from it.
3M.F. Craun, et al., Waterborne outbreaks reported in the United States, J. Wat. Health 4(Suppl. 2), 19–30, 2006.