What Is The Value of AI/Big Data for Water Utilities?

It’s getting worse!

We won’t belabor the point, we have a growing water quality issue in the US. The NYT reports issues in 1000 SF cities, Newark is looking worse than flint, and Chicago. Well, Chicago. In Q2 2019 87M people were exposed to poor drinking water with 160K Violations! And we believe that’s an undercount.

A utility operator for a west coast city shared with us that they dump 6M gallons of water a week because they don’t know whether it all flowed through the system or is stale. That’s both unsustainable for the environment and financially for the utility. When we dug further, we found that the regular culprits — crumbling infrastructure, water source depletion — were present. But there was an unspoken issue of information not flowing across the organization at the speed and fluidity (no pun intended) that is required to ensure operations run smoothly. There were operational silos that prevent adequate responses when issues arise — the right hand did not know what the left hand was doing (operations did not know what the chemistry team was doing. And neither the chemistry nor operations team knew what the financial implications of what this one single action of dumping all this water was. The teams knew it was not ideal, but couldn’t put a financial number on it (nor think in those terms) and (pun intended) flushed money down the drain.

Tier 2 city utilities are suffering the most.

The examples above are all tier 1 cities. And these are the cities who have millions of dollars to work with expensive consulting firms or pay for multi-million dollar software products. But as the list of incidents at the start of this article shows, even those tier 1 cities are struggling with the operational issues. They are getting better at avoiding preventing the water quality issues that arise from poor system management and are stemming the losses that happen as a result of the operational waste. So it should come as no surprise that the tier 2 utilities (with customers between 10k-200k) are struggling even more. 6k of them had serious water quality violations at least once in the last 12 months, some with multiple violations. And again, these issues are under-reported.

What’s needed?

Collection of data, analysis of that data, and recommendations to enable the manager of the utility to make the right decision at the right time to save money and time. That’s what is required to improve water utility operations and prevent water quality issues. Of course, we also need new infrastructure and employee training. But the new infrastructure cannot be laid over old operational approaches. The new data-driven approach to running the utility will bring about a need for new processes.

Tier 2 city water utilities do not have the information and insights required to save time and money on addressing issues and running their operations. Information about what is going on and where an issue is/might be. Information shared across the organization, across silos, in a seamless way so that the whole team knows what’s going on and who is responsible for what. And insights. Insights obtained from the analysis of that information to determine how to optimally address the issue.

There is a lot of information out there relating to the water utility. Troves of data that can be mined, in our case using AI, to convert that data into insights. To increase the fidelity of the data, sensors within the system can be added to;

1. Make information useful about the what, where, why and what to do when it is needed (essentially, situational awareness) relating to a quality issue
2. Predict what will happen to help the team better prepare: a utility manager told us that before noon his plans for the day have already had to change due to changes in water system quality or tank levels.
3. Most importantly, put information and insights in the hands of who needs them especially as the old cadre of utility operators are retiring and the new crop have to ramp up quickly.

How will AI/Big Data benefit these utilities?

There is a lot of information out there relating to the water utility getting you and I clean water every day. Troves of data that can be mined, in our case using AI, to convert that data into insights. To increase the fidelity of the data, sensors within the system to capture more data. Using sensors and AI (as Varuna provides) enables the utility to detect and address issues earlier than previously could, reducing time and cost. For example, the 10-day zebra mussel issue in Austin could have been reduced to 6 days as the first few days were spent trying to identify the source of the issue! Using AI or big data analysis, these tier 2 cities can see the operational costs and times from other towns addressing the same issues and consequently save themselves time and cost on contaminants or through optimal resource allocation. Data-driven decision making improves service quality, a metric measured through customer satisfaction. It’s proven out in other industries and can be applied in water utilities.

Who is doing something about it?

It’s shocking how little interest is in this space by startups (understandable) due to the unwillingness of the utilities to admit there is something wrong; the first thing we always hear during conversations is ‘our water is fine’.Only when we’ve gotten them to see we are on their side do they share the issues they have. And we are solving those issues. With a full system that collects information (on the where, why, what), analyzes that information (to provide the what to do and by whom) and provides them prognostics that help with better planning for predicted events. There is a trove of data out there on financial state (how much did it cost this utility to fix this problem?), weather data, source water data, etc. and when you mine this trove of data and lay it over the situation of a tier 2 utility, the value is immense! Especially for these utilities, most of them, who do not have the operational or data analytics/AI capabilities to analyze the data and draw these insights themselves. They are working hard but it’s time for them to work with a partner who cares.

Conclusion: In some industries, the lack of information is a nuisance, things keep on working nonetheless. But not in the water industry. 87M Americans experienced a violation in the last quarter. Lack of information is literally going to kill us if we don’t do something. At Varuna we’re doing something, -we’re helping water utilities deliver good quality water by removing operational waste — reach out to work with us as a utility or a solution provider (we’ve even built APIs to make this easier for all involved to share the data that is required and integrate easily!). Join us! Or get your local utility to join us!

Varuna helps water utilities deliver clean water by removing operational waste. Sign up for Designing H2O, our weekly newsletter.