Dr. Peter Williams opens EWTS with keynote address

ONTARIO, CA -- Dr. Peter Williams opened the sixth Emerging Water Technology Symposium with a keynote address that examined three emerging technology trends -- IoT, Big Data and Artificial Intelligence -- and the four industrial and social transformations he believes they will produce.

The sixth EWTS is co-convened by the four associations that have worked together closely over the past few years to grow the symposium: The Alliance for Water Efficiency (AWE), The American Society of Plumbing Engineers (ASPE), The International Association of Plumbing and Mechanical Officials (IAPMO), and Plumbing Manufacturers International (PMI), in cooperation with the World Plumbing Council and the Symposium's industry and media partners. 

After opening remarks and a welcome to attendees from Russ Chaney IAPMO CEO, Kerry Stackpole, PMI CEO/Executive Director, Billy Smith ASPE Executive Director/CEO and Mary Ann Dickenson, AWE President/CEO, Williams took the podium.

Williams has played a major role in the development and delivery of IBM's smart cities, water management and resilience solutions. He co-wrote the UN's Disaster Resilience Scorecard, used by about 40 cities globally, and scheduled to be used by a further 200. A native of the UK and a resident of the US, he holds a doctorate from the University of Bath, the title of IBM Distinguished Engineer, and is a Visiting Lecturer on Smart Cities and Communities at Stanford University.

Williams began his keynote by looking at three emerging technology trends, IoT (in particular Smart Cities), Big Data and Artificial Intelligence, and looking to four transformations they are initiating.

"Change may come slower to the water industry than some other industries, but it is coming all the same."

First, large scales are becoming small scales. Big Data, he thinks, should actually be called Small Data, because it enables us to see on a finer, smaller, more exact scale the world around us. Instead of daily or hourly measurements we are moving towards continuous, real-time measurements.

As an illustrative example, he cited Thames Water, a UK utility looking to improve water efficiencies by examining where leaks were happening. Instead of tracking data from postal codes or even city blocks, the company set up sensors on more than a million individual manholes. They discovered that less than one percent of the sites were responsible for more than 25 percent of the leaks, and were able to develop cost-effective strategies and predictive models that gained significant efficiencies.

Second, stand-alone systems are becoming integrated. More and more systems are leveraging connectivity to improve performance. For example, energy utilities with their demand response.

Third, reactive is becoming proactive. Instead of identifying the problem and fixing it, the new model will be predictive maintenance; fixing the problem before it ever happens. It means, "More uptime for the same maintenance expense," Williams said. This proactive model extends beyond the simple, "Which pipes might be about to burst?" all the way to revenue prediction, as in "Which customers may have trouble paying their bill?"

And fourth, analytic will become self-learning. Most people -- especially marketing people -- when they refer to Artificial Intelligence, are simply talking about analytics. True A.I. will incorporate structured and unstructured data -- both what operators deliberately feed it, and what it can interpolate from raw sensor data. True A.I. will be able to detect patterns and issues that would escape a human being. A.I. systems will be learning systems, meaning they will get better -- provide better predictions, generate more useful suggestions -- the more they are used.

Of course, all these transformations come with attendant risks. Integrated systems are vulnerable to cascading failures. One tree branch falling on a power line in Ohio managed to black out a swath of North America extending up to New Brunswick in Canada back in 2003. And the scary part? Almost every part of the system was only doing exactly what it was designed to do.

There are emergent privacy threats caused by individual data collected by smart systems. Sometimes innocuous data points when carefully connected can produce a privacy threat. That data, however, is going to be collected just the same. The benefits to be gained are simply too great. Instead of trying to prevent the collection, rules and legislation need to be put in place regarding its uses. And in the meanwhile there are cyber criminals, working often at the limits of technological innovation, where safeguards against its misuse have yet to be developed.

We live in "A world that is changing profoundly," Williams said. "Change may come slower to the water industry than some other industries, but it is coming all the same."