Will big data lead to a dumb future for smart water?

Published August 4th, 2016

Cg head

Insight from Christopher Gasson, GWI publisher

A few weeks back I met up with Veolia’s Laurent Auguste in London. I have always rated him as one of our industry’s great visionaries, and we were talking about machine learning in the context of the impact of big data on the water industry. “It is scary,” he remarked. “One day they might put us out of business.”

On the face of it this statement seems absolutely wrong. Big data is an important driver of business for an outsourcing company like Veolia. The move toward smart systems can be traumatic for public utilities: it requires a level of IT expertise and infrastructure that few have in-house, and even if projects can be implemented effectively it often leaves managers overwhelmed by data rather than empowered by it. Veolia has a joint venture with IBM to deliver “digital urban solutions” and it has the scale to be able to develop smart water solutions in a way that no stand-alone utility could match. Suez too has put smart water at the centre of its outsourced services strategy, and seems to be delivering excellent results from that business line.

So what can artificial intelligence do with big data in water that is so scary?

Hitherto we have thought about smart systems in terms of better billing, better customer service, better asset management, better water quality control and better water resources management. There is software and data management solutions for all of these. Veolia and its competitors like Suez can make a margin by knowing how to implement these software packages and the related monitoring and control systems. The scary thing is that once there are data-feeds governing all the major aspects of utility management then there is a level playing field between men and machines. Instead of a manager with years of experience being best placed to make decisions about how to optimise overall operations, cognitive computing makes it possible for a machine to become the expert. All the important decisions could be made by something like IBM’s Watson question-answering system backed by hyperscale cloud-based computing.

The humans meanwhile would be relegated to jobs like digging up the roads and hacking fatbergs out of sewers.

GWI is publishing a report on big data in water next month. It doesn’t go as far as Auguste’s vision, but it does show that data-driven monitoring and control systems are changing the face of the water industry. We are also arranging a workshop on the subject with the Global Water Leaders Group in Miami in December. Both should be fascinating. For further details contact Jake Gomme on jg@globalwaterintel.com.