Data security is a primary concern for the U.S. In May 2013, the U.S. Government awarded Amazon a security clearance to allow federal agencies to easily use its cloud computing services as part of the Federal Risk and Authorization Management Program (FedRAMP). FedRAMP is a mandatory government-wide program that standardizes security assessment, authorization, and monitoring for cloud products and services. The Program’s Cloud First policy mandates that agencies take full advantage of cloud computing benefits to maximize capacity utilization, improve IT flexibility and responsiveness, and minimize cost.

Like the U.S. government, TaKaDu uses Amazon Web Services (AWS) for its cloud software system. Amazon’s multi-layer security is built on world-class physical security, environmental safeguards, IT security, business practices and transparent auditing, as well as design for high availability and backup. As a global leader in Water Network Monitoring, TaKaDu provides Software-as-a-Service solution for its customers. All network processing is performed through Amazon’s remote EC2 (Elastic Compute Cloud) computing facilities. Amazon’s automatic scalability allows users to define specific conditions such as hourly, daily, or weekly variability in usage. TaKaDu’s exchange protocols and data stream encryption also meet or exceed leading industry and government standards.

According to TaKaDu’s VP of R&D, Benny Keinan, “A software company’s most important asset is its ‘source code’, which holds all the ‘secrets’. Recently, there is a trend of software companies shifting from an on-premise source control solution, to use cloud services for source control such as GitHub or Amazon. IT security is not TaKaDu’s core competency, thus it makes most sense to allow an outside, expert provider such as Amazon manage its secure data instead of building its own data center. This allows us to provide our customers with the highest reliability and security level.”

Leaks and bursts appear to be a fact of life for water network operators. Pipes, joints, and other parts spring new leaks under the normal wear and tear water network  infrastructure is subjected to. The question is: how do you deal with leaks and bursts? Can you delay their onset? How soon can you find and repair them?

Water loss control, as its name implies, strives not so much to eradicate this phenomenon, as to restrain it. For example, for a constant rate of occurrence of new leaks, the amount of water lost depends simply on the time it takes for operations staff to find and repair those leaks, so earlier detection and repair means less Non-Revenue Water. Incidentally, it also means less leaks have time to develop and escalate into large, visible bursts and service interruptions, so keeping water loss under control may improve several distinct measures or KPIs.

So, what is the “magic tool” for water loss control? Some products are marketed as “THE tool to stop water loss”, but I’m not sure that’s possible. I always like to say that it’s a problem which takes a toolbox, not just one tool, and just recently I ran into this fantastic chart presented by a great customer of ours, Yarra Valley Water (from Melbourne, Australia), which I think demonstrates this very neatly. It’s from a slide in this very interesting presentation by Ken Thompson. This graph shows YVW’s total Non-Revenue Water (dark vertical bars) and estimated leakage per connection per day (orange line) over 16 years, with an overlay of the water loss control measures put into use during those years (captioned grey arrows, running from the year each system or method was deployed). In percentage terms, that last bar is somewhere under 12% NRW.

I can’t help saying that Yarra Valley still reduced  about one fifth of their remaining annual loss rate when they deployed TaKaDu, although they managed to reduce water loss considerably over the years.

It’s an amazing journey, and one has to applaud Yarra Valley Water not just for persevering and succeeding so spectacularly, but for managing to zoom all the way out and see the “big big picture” of this process, and then finally for sharing it so clearly.

There is a lot to be learned from this graph, but here are my five quick takeaways:

1. Non-Revenue Water (and leakage, specifically) is clearly influenced by technology and methodology. It is no exaggeration to state that the nine tools shown here have revolutionised water loss control at Yarra Valley Water, leading to a four-fold improvement. Water loss is clearly not some insurmountable “force of nature”. Interestingly, the reduction in water loss has been economically beneficial for YVW, as changing methods and systems have in effect driven down the company’s Economic Level of Leakage.
2. Getting the “low-hanging fruit” can have a dramatic effect. Improved metering in the first  few years of this water loss control program was responsible for most of the reductions achieved to date.
3. Conversely, even smart, well-performing network operators can (and should) still improve. It’s not always easy to decide how to attribute the impressive reduction in water loss to the various solutions in use, and some benefits may only play out fully after some years of practice and process integration. But the continued downward trend is unmistakable, as is the immediate contribution (abrupt year-on-year decrease) when certain solutions were introduced, most lately TaKaDu. A number of individual measures each seem to have cut 10-30% off the existing water loss numbers. Cutting total NRW and leakage by about 20% at an already well-performing water company – the latest step in this chart – is an impressive improvement, and I believe Yarra Valley Water will continue to drive water loss further down, both through realising more completely the benefits of deployed solutions and through continued adoption of cost-effective technologies.
4. “Bad years” do happen, but a well-directed multi-year campaign to reduce water loss can show consistent improvement almost every year.
5. There is no magic bullet to individually stop water loss. However, there is quite an array of available tools to add to the water loss control toolbox, some of them effective even after the easy problems are fixed. That’s not so bad: six “small” step-changes, each improving  just 20% on previous achievements, add up to a 75% reduction, and based on YVW’s example, could be deployed effectively at a rate of one every year or two.

I didn’t get to beautiful Utrecht to give this talk on SWANonomics at SWAN 2012, as I had intended, although I did get a great substitute to stand in for me.

Rather than talk about TaKaDu, network monitoring, or any of our technology, I wanted to use our few years’ experience in this dynamic part of the water industry to shine a light on some of the more surprising economic aspects of the Smart Water Networks revolution. Data revolutions are funny that way. It’s not just the tired truism that adoption of any technology is an economic process; new data has a way of changing what you know about the real world and your existing processes, and creates real new opportunities to take action differently.

Here are some of the unexpected headlines:

-          Good news: smart monitoring drives spending increase on field crews

-          Accepting new values discovered to entrenched KPIs goes through the exact 5 stages of grief

-          Better leakage reduction means less Ml/day repaired

-          The cost of water not lost is just a small part of the value of water loss reduction

Miriam Alster, Flash 90

By Amir Peleg

Last week, Some of TaKaDu’s employees and I had the pleasure of meeting Israel’s President, Mr. Shimon Peres. We met the president just as he was returning from the World Economic Forum at Davos, where I happened to meet him for the first time, two years ago, when TaKaDu was recognized as a World Economic Forum Technology Pioneer for 2011.

This meeting was at his request: he wanted to learn more about how innovation can solve some of the world’s water problems and the looming water security crisis, recently ranked high by the Global Risk Survey 2013 of the World Economic Forum.

President Peres commented that the water crisis is one of the more imminent global crises facing humanity. We spoke about the fact that over 25% of potable water in the world is lost in the water distribution system as a result of leaks, bursts and other network issues (the professional term is “non-revenue water”). I was impressed by Mr. Peres’s interest in water efficiency and its wide implications, and his excitement about the opportunity to impact global problems through the use of new water technologies.

Miriam Alster, Flash 90

We presented TaKaDu’s vision of water network monitoring, to reduce water loss and improve operational efficiency by identifying and alerting upon network issues as they occur, through data analytics. I made the point that the issue is not as simple as it seems. The scarcity of water as a resource is not the only issue at hand. The bigger scarcity in the water sector is scarcity of technology innovation. This innovation has the potential of resolving many water sector issues. Israel’s high tech industry can serve as an optimal breeding ground for water technology innovation, since it has the three major requirements needed to foster technology innovation: (1) an entrepreneurial spirit and seasoned entrepreneurs, (2) a high degree of awareness of water issues, and (3) a well-developed venture capital community and government support of R&D efforts. Water innovation can take today’s water networks into the 21st century, solving water scarcity issues by better management of water distribution networks.

Mr. Peres expressed his support and hope that more people, in both industry and academia, will see water innovation as the core theme of the companies they found and the technologies they seek to develop.

The global risk that respondents rated most likely to manifest over the next 10 years is severe income disparity, while the risk rated as having the highest impact if it were to manifest is major systemic financial failure. There are also two risks appearing in the top five of both impact and likelihood – chronic fiscal imbalances and water supply crisis. Here is the report’s figure 4 ranking risks by likelihood and impact.

I went to Water Loss 2012 in Manila to tell water utilities one simple thing: data already collected under accepted “best practices” is all they need for a dramatic improvement to water loss control. Of course, that’s part of what you get when you deploy TaKaDu, but I wanted to focus on something else.

There are various tools and methodologies out there to use network monitoring data, and they are far from having been created equal. Whilst the data from sensors and other sources does hold the key to water loss control, and the principles as sketched on a conference slide are simple enough, real-world conditions make data analysis the trickiest link in the active leakage control chain. Three murky clouds – typically glossed over in water loss presentations – help to muddy the waters:

-          Data quality (the meter values could be wrong)

-          Other network events (the event you found may not be a leak)

-          Complex utility process (you think there may be a small leak somewhere – what to do now?)

In this presentation (and the more detailed paper), I revisited the “traditional” Active Leakage Control process, highlighting the role of data analysis (manual and automated), revolving around leakage analyst’s work on the 3 “S”es of data analysis Supermen: Sifting, Statistical estimation, and Special knowledge. By boosting this difficult stage, utilities report they achieved significant quantifiable savings throughout the ALC process.

To do this, I listed some guidelines for coping with the 3 big uncertainties of data, network events, and the utility process, starting with detailed knowledge and understanding of these factors, to be addressed by suitable processing. As with many data analytic challenges, these real-world data “technicalities” are, in fact, the main challenge for data-driven ALC.

If you’re a pessimist, my slides are mostly a long list of impediments to active leakage control. If you’re an optimist, they are a collection of opportunities to do it better. If you work at TaKaDu, these slides are just what we have been doing for the past few years, what we’re good at doing, and what we need to keep doing. Have a look also at some of the other great talks and papers from the conference, too.

The OECD and the European Commission held a fascinating workshop earlier this year in Copenhagen titled “The future of Eco-Innovation: The role of business models in green transformation”, and I was fortunate enough to be invited to speak there. A handful of representatives from some truly amazing companies were asked to help the workshop experts “showcase good practices of radical and systemic eco-innovations”. What a great validation for our belief in TaKaDu’s potential to transform the water industry!

It’s not an easy subject for idealists, policy-makers, academics, and people from industry – all present at the workshop – to discuss, though. Whatever their beliefs and private motivations, business leaders need to build viable businesses, where “green” technology can triumph only if it promotes great tangible value. This can typically be through marketing and public perception, concrete cost savings, or regulatory compliance. Yet on a national or global level, innovation for sustainability is often recognised as a goal in its own right. Can the two paradigms be brought together, and not just at the superficial “greenwashing” of entrenched industrial methods? Are big government grants for R&D consortia the best way to encourage this – vying with business investors to steer the industry, or should policy-makers wield their power in more unique ways?
For the case study of TaKaDu, I tried to explain what makes our technology, business, and industry so special, and how they are inextricably linked. Building on that, I shared some views on where government or international economic policy (the OECD’s main interests) can encourage and shape ecologically beneficial innovation. This, we hope, can harness and reinforce the entrepreneurial spirit and innovative capacity of business and technology innovators towards globally beneficial goals.

To continue ensuring that Jerusalem can continue its thousands of years sustainable water use, Jerusalem’s Hagihon utility is one of the first utilities to have deployed a smart water network using TaKaDu’s water network monitoring solution.

Earlier this week Hagihon and TaKaDu were the joint hosts of a local industry event with Israel’s water utility managers. The event featured a case study of Hagihon’s use of TaKaDu and presented Hagihon’s results: detecting hidden leaks, energy savings, identifying meter issues, water quality monitoring etc.

Hagihon’s general manager, Mr. Zohar Yinon, emphasized that Smart Water technologies are used to improve Hagihon’s overall operations, with an emphasis on maintenance prioritization and customer service improvement by allowing Hagihon to proactively resolve network issues, such as hidden leaks.

Mr. Oded Distel, head of Israel NewTech, who was a key speaker, highlighted the importance of sharing Israel’s water utilities’ expertise with utilities in other countries.

Amir Peleg, TaKaDu’s founder and CEO said that there is no replacement for the continuous dialogue between technology companies and water utilities, to define and resolve the main product and technology needs for the future’s utilities. Amir also thanked Hagihon for its early deployment of the TaKaDu system.

I’ve been working with Hydraulic modeling applications for more than 15 years. Ever since I began working with TaKaDu, a water network monitoring pioneer, I’ve been asked whether hydraulic modeling and water network monitoring collide or are complementary. In short, I will argue that one solution (good network monitoring) can make the other solution (hydraulic modeling) significantly better. Explaining how is the purpose of this blog post.

Hydraulic modeling: medium and long term analysis

Hydraulic modeling is a technology that has been used in the water utility space for more than 30 years. It is used to answer the big questions, medium and long term, on a water utility’s planning horizon:

• Will a bigger pumping station be needed? Where?
• A larger pipe?
• How will the system behave upon a major growth corridors expansion?
• What infrastructure is needed twenty years from now?

The key strength of hydraulic modeling applications is the ability to work with “what-if?” scenarios. For instance, in the case of flooding – what if a pump was turned off or a certain area needs to be evacuated?

Moving into operational models

In the recent years, there has been a focus on the potential use of hydraulic modeling to answer short term forecasting / prediction questions, specifically in the water distribution network. It could be used as a decision support tool for operational network managers.

There is one problem: when the utility’s telemetry, operations and modeling staff meet to discuss operational hydraulic modeling, they are typically nervous. A consistent list of concerns repeatedly comes up:

• Good model outputs require good data coming in and telemetry inputs are notoriously fickle – they are not consistent or accurate enough for this application.
• Good model outputs require an up to date model. Yet there are high costs in keeping hydraulic models updated and new infrastructure is added to water systems day after day. Models are generally only updated every few years and are not up to date.
• Good model outputs require a good model of how water is consumed by customers – what the peaks and troughs are, how they vary between users, neighborhoods, the seasons and days of the week.

How water network monitoring fits in

Water network monitoring has the potential of complementing operational hydraulic models. For instance, water network monitoring can significantly assist in solving the telemetry quality issue. To ensure that a consistent high quality feed goes into the hydraulic model, TaKaDu can provide inputs on data quality and assist the utility in making the telemetry data better.

The reason is that good water network monitoring will audit and clean inconsistent telemetry data.

Another example is using water network monitoring to obtain high quality data on how water is consumed throughput the system by day, date, neighborhood and other criteria.

All in all, most utilities aspire to  have high quality hydraulic operational modeling systems as an important tool for critical decisions. But to seriously use them they must first address the data quality issue, the question of how is the network operates today before they try to predict the future.

This is how water network monitoring makes operational hydraulic models better.

When I was asked to speak at the Intelligent Cities Expo in Hamburg about Smart Water Networks, I assumed I would be on a double defensive: apologizing for an industry late to catch up on the data revolutions of the past hundred years, and struggling to demonstrate the value of data in water against a backdrop of data-poor, “conventional” engineering water innovation. I’m glad to say I was completely wrong!

At the water-themed session I was flanked by brilliant colleagues Steffen Schaefer from IBM and Francis Campan from Suez Environnement, talking data-driven solutions from start to finish. With these two industry giants helping to drive home the point (one from the data processing side, and one from the water business), the clear message was that smart water management is data-driven water management. We were practically “singing from the same hymn book”, perhaps not surprising, considering our three companies’ commitment to SWAN – Smart Water Networks Forum. Moreover, contrasting the session as impartially as I could with other sessions on power, heating, transportation, and other city infrastructures, water seemed to stand out as the industry with the clearest “program” (however unofficial) to get the right data and put it to good use.

In this presentation, I defined the “Smart Water Network Revolution” and the benefits it is bringing, talked about how policy-makers and water professionals can and should help it along, and did my best to dispel the four great myths about water utilities and adoption of data technology:

1. Water utilities are too conservative or low-tech to adopt such technology
2. Water is a low-value commodity, therefore not a magnet for industrial R&D innovation
3. Water utilities are not yet ready for “smart data systems”
4. It’s too big an investment

Flip through the slides for the reality behind these myths.