The Dawn of the Invisible Water Utility

hi i'm peter prevos i'm the manager of data science at columbia water which is a water utility in regional victoria here in australia and it's an honor to be with you today i'll all be virtually at the water show africa and today i'm going to talk to you about the dawn of the digital water utility so i want to talk to you about digital transformation and the role that plays within water utilities but before i get into a case study and give you some advice on how to implement digital transformation i'm going to provide a little bit of context when i grew up in the 1970s i used to love watching shows about the future about everything past the year 2000 and back in those days in these documentaries people thought that we would be living in a utopia and that technology would be the hero that created the world without poverty and without disease we would all be flying around in cars and for some reason the fashion of the future would be that we had aluminium onesies now there's another story of the future and that is in science fiction the science fiction stories that i used to love to read and watch the movies they paint a very different picture it's not the utopia of technology of the documentaries but it's more a dystopia where technology is the villain instead of the hero where technology has created a world of environmental destruction that is full of disease and poverty and instead of wearing aluminium shiny suits would be wearing hashing bags or something like that i think we now in the future that's sort of halfway in between perhaps these two extremes that technology has developed a lot and some of the things that we are capable of doing now would only would be pure science fiction only 30 40 years ago but there's also a slight dystopian aspect about our vision of the futures because we know that climate change is a real issue and especially with our industry that this is a problem that requires a lot of solutions now south african president cyril ramposa said that unless we take drastic measures to conserve water sources and promote efficient use water insecurity will become the biggest develop developmental and economic challenge facing this country and i think we can extend this worst to say that it's not only south africa but it's the whole of africa and even the whole of the world now waterline technology are really very very old partners and water management has always been based on technology because we have filtration and pumping and biological sciences we always very science based and one of the very earliest examples of data science ever is actually in the water industry where dr jon snow started mapping all the cases of cholera within within a certain area of london and by doing informal spatial analysis he worked out that it was a particular pump that was causing a problem and this created the insight that water is actually a cause of disease and that was a revolutionary technology so this image here shows a modern analysis of the data that was collected by jon snow but what has happened things have changed a lot since the days of jon snow when everything was on paper and analysis had to be done by slight wall or by hand so what is this digital revolution what is this data revolution that people are talking about well the first part of the revolution is that a lot more data is currently collected now collecting data is as old as humanity and as a matter of fact the oldest data record in the world ever was found in what is now the democratic republic of congo and it's the ushangu bom which is now housed in a museum in in belgium and it's a a 10 centimeter long bone of a chimpanzee and it has very distinct markings on it and that was a way for those people to record data now scientists are not quite sure what it actually means whether it's a calendar or whether even it's a sort of a device to do calculations but the point is a data storage has been around forever and a day and all through history people have been collecting data but what has changed now is that everything most data collection is now electronic and electronic data has great advantages because we can collect so much more of it collecting it is much cheaper and it's also much easier to find your way within that data so all processes that we have as businesses they leave a digital footprint so if the customer rings up and they leave a complaint we have a digital trail of the whole conversation that is have with our customer and and how the problem was resolved we have the scada system that does a lot of measurements and and creates a lot of data that is used to control the plant we have laboratory data we're building down the meteorites and so on and so on all the data is used for a specific purpose to manage the process here and now and then it's stored in databases and usually it becomes what some people call dark data and they get stashed away now a lot of my job is about using that data and recycling it to create new value so the first part of the data evolution a lot more data is available the second part is that computer speed and memory capacity has exploded over the last 50 years almost every 18 months over the last few decades the capacity of computers has doubled in 1985 when i was a teenager and i got really interested in computers i was salivating at pictures of the cray 2 super computer which i'm showing here on the screen and this computer which is the size of a small of a bathroom at that time was the fastest computer in the world now think about the fact that this computer has the same capabilities as an ipad 2. so in other words we now have super computers within our pockets so we have a lot more data available but we also have the tools to store and process that data which makes it much easier the third part of that revolution is that we have a lot more tools available to analyze that data and especially through open source software so open source software itself a software where the recipe is available and it's shared between people while some some might think that is not commercially smart it actually accelerates innovation and open source tools such as the python and the r computing languages have accelerated what we can do with with data and more about that at the end of my talk so we have more dial available we have the machinery to process it and we have the tools to to analyze the data and visualize it and make it useful and this has created this data revolution but the question is how can award utility maximize the benefits that are available from this technology what can you do to squeeze more value out of this information now the term digital wall utility i think is slightly confusing because many people talk about the digital water utility as if the industry will be digitally disrupted but i don't think that's technically correct what utilities will never be digitally disrupted we can only benefit from digital technology because digital disruption occurs when the offering of a service provider is replaced with a digital offering so you might have a brick and mortar shop and then your competitor has a webshop and then you are digitally disrupted you might have a hotel booking system where people ring up and and make hotel bookings but then a website comes along and digitally disrupts your offering travel agents and so on and so on water utilities will never be digitally disrupted because we don't have a situation where ones and zeros come under a tap instead of water water would always be a physical service that weighs a ton per cubic meter and has to be pushed uphill but we can't change these laws of physics but what we can change is by having better information by analyzing it better is being more efficient and how we manage that water resource so to define the benefits of digital transformation we need to know what is a perfect water utility and the value of of a service and the value of a border service is very simply defined as the difference between the benefits and the cost now traditionally when we talk about the cost of water we talk about the monetary cost the amount of money that people pay to get their service but i think that most important aspect of a water service is the time cost that customers pay for this and pay between quotation marks so the time cost in a developed area with a very well functioning water service is almost zero and it approximates zero because all you have to do is you open a tap water comes out and you never have to do anything else the water comes out it's available it's clean to drink you don't get sick if you live in an area that doesn't have improved water surfaces you might have to spend several hours a day to carry the water to where you need it and if you get sick you incur a large time cost because illness or public health is expressed in disability adjusted life years so in other words being sick is also a time cost and the perfect world utility is one that is almost invisible to the customer that we reduce the time cost to approximate zero so that the water is always available and it's always safe to drink come to think of it reliability is the greatest benefit that award utility can provide now the famous taxi company uber their strategy a few years ago was literally worded as that they want to be as reliable as running water so in other words a company like uber that you know probably the example for a lot of organizations around the world wants to be as cool as a water utility so just think about that for a moment but this invisibility comes at a price there is a paradox in being invisible because if customers don't know who you are then also you are not able to build a relationship with your customer you still need to minimize the time cost talking about your core services but you need to also build a relationship with your customers now technology can actually help you to become an invisible water utility with a strong relationship you can use technology and you can use data analysis to create a water service with a time cost that approximates zero because we can use technology to find problems before they occur and then fix them before customers or the natural environment for the matter even notices that a problem occurs but this same technology can also be used to build relationship with customers because the technology allows us to communicate with them as well and if you are interested um in this concept of the invisible wall utility up um i'll do now a shameless book plug you can read more about this in my book customer experience management for water utilities which is available from iwa publishing and that's the commercial message on the next part of this talk i'm going to talk to you about a project that i instigated within my own water utilities where we are converting all our existing customer water meters to be digi with two digital meters but before i get into the technology i want to start explaining why we actually embarked on this journey so we didn't sit around the table one day and decided that we needed digital transformation we sat around the table and we started discussing problems so i was speaking to different people in the organization and i heard different messages about issues that they were experiencing now the revenue team started talking about the issues that they're experiencing with managing our major physical meter reader activities so we have about 80 000 meters spread over over a fairly large area and we have four meter readers and each of these meter readers has to walk about 1500 kilometers per year to read every single water meters four times now there's a lot of effort involved in that there's also oceans issues because australia being famous for its snakes and spiders are meter readers occasionally have to encounter these creatures from other parts of the business i heard that it's very hard to actually do what they need to do and to analyze data because with only four data points available from each customer it's very hard to balance the books it's like trying to figure out where your money went by only looking at your bank account four times per year it is impossible to to really do what you need to do and because it is limited data availability all our investment decisions were based on a lot of assumptions because we simply didn't have enough data available about custom consumption and the other aspect is that we were making estimates about non-revenue water so we know the leakage in our system we've we we think it was a fairly reasonable amount of leakage but again we had to rely on a lot of assumptions so when those three problems came together around the table that's when the decision made was made to start tendering for a digital solution now i call them digital water meters because some people call them smart meters but i don't think the word smart meter is quite justified because what we do all this device does and here's an image of the the technology that we purchased from a company named ventia this company of vino within within the larger company via invent here and this device plugs in into an existing meter so we don't actually replace the meter but we put a device on top of the meter and it gives us a signal every hour 24 hours a day so in other words we're going from four meter reads per customer per year to about two and a half thousand meter reads per customer per year so that's an enormous increase in in the in the data and it also means that we don't have to send meter readers out anymore or a lot less um to to do their activities so how do we achieve those benefits what is that what does that look like now the first thing we look at is technology effectiveness so we get a lot of data and we've developed our own databases and our own analytics to look at this information and here's a picture of a town called rochester which is just north of bengal and there are about a thousand connections in in rochester and this is one of the charts that our contractor venti has produced but we produce similar charts where we look at the amount of data that's actually coming back from these devices and blue means that we get a lot of data and i think it's defined as at least 15 per day and you see there's a few customers there that we didn't get a lot of data from and this chart then helped us to redefine the network to make sure that we get more data so this the first part of our reporting is just looking at the technology effectiveness and the second thing that we are now able to do is and do some highly detailed network monitoring and the image i'm showing on the screen now is a diurnal curve of one of our towns where we look at the water consumption each hour of the day by adding up all the data we get from the individual customers so this tells us exactly how much water is consumed at each hour of the day and we we can also do this per customer segment and i could look at specific streets or i could look at commercial properties or something's about large gardens or customers with our gardens we can now start to fine-tune this data and learn a lot more about the consumption of our customers and what we can see in this graph here is that in this particular town there is and it's not a very large town so the numbers are not that huge but there is between midnight and six o'clock in the morning there's a fairly bit of a fair amount of water going through the system and there is no other reason to explain that than that this is leakage because there's not much nightlife in these small towns in regional victoria so we have now a lot more information and we can now use this to benchmark our lead detection process and see if we can bring this curve further down now the technology itself only gives you data that allows you to make a decision to invest in lead detection and in repairs obviously the technology itself does not actually fix the leaks for you now here's a another example with that same data which is non-revenue water before we had this technology we were only able to measure non-revenue water four times per year and we had to estimate it and it was quite a big process to measure the amount of non-revenue water but this orange line here shows the non-revenue war on a day-by-day basis or you can even zoom in on an hour by hour basis because that's the granularity of the data we receive and you see here that we found this mysterious spike in non-revenue water that went for a few days in this in this town of trenton the same time you you saw previously and this data though helps us to decide whether we whether this this issue with non-revenue order is big enough for us to spend some money on trying to find out where it is because unfortunately this this technology does not allow me to find out where it is unless i have some more zone meters perhaps so this is this is all these are all the benefits i guess you could say on the business side of the equation where we are able to optimize our systems but there's also a very strong customer aspect to this project so going back to the invisible utility we can now manage our systems better and perhaps we can reduce the time cost for our customers but we can also help with this project to build better relationships and the very first thing we did when we decided to embark on this journey is to start talking to our customers and this image here you see my colleagues stuart and darren and we were here in the town called wetterburn to talk to the local community about this technology and to learn to ask questions and to start to promote the idea now one of the things we can do now is that when customers have a high bill and they ring us and they already have a digital device attached to their water meter we can all have much more meaningful conversations and our staff are able to pull up a dashboard like what i'm showing on the screen now and on this dashboard here you can see and this is a real life case where a customer had a very high water bill they rang uh customer service staff and by talking to them they found out that they had an issue on their property part of their sprinkler network had been going for a very long time and after talking to us and we talking through the problem they were able to switch it off and their bills were now a lot lower so this dashboard is one that's not necessarily directly available to customers but customers who have a problem they can talk to us and our customer service staff have been trained in interpreting this information and they can now have meaningful conversations with our customers so yes we will the technology allows us to be more invisible but it also allows us to build better relationships with customers now the next step we haven't quite done this yet is to actually provide customers the ability to self-serve and find this information although i'm not a strong believer that we should necessarily provide customers with detailed raw data and graphs because not everybody might be equally interested or capable of interpreting that information so the ideas we're currently working on is to provide customers with push notifications and where we can for example sending a message that we believe you have a leak on your property because lead detection online property is quite easy to do mathematically um we are thinking about some gamification perhaps where we can do social comparison and tell customers that they are using more or less water than everybody else on their street and there's lots of different things we can do to provide additional value to customers to help them manage their water use better now to wrap up this talk i'd like to go through a few points of what i've learned in digital transformation strategy and some advice on how i believe he should proceed and i think the most important part is when you talk about digital transformation is don't focus on the technology first define the problem that you want to solve so as i explained in our digital water metering project we had some problems some areas around the business we put them together on the table and then the solution became apparent that we wanted more data from our water meters so to find a problem and then find the technology that sort that will most likely help you to solve that problem and then when you searching for that technology don't try to specify everything to the nth degree so when we did our digital metering projects we did not specify anything about transmission frequencies and all these sort of technical things that we are not really experts on we move that risk into the contract so that we can focus on deriving benefits from the dialogue because that's in the end what it's all about the other thing i would strongly advise is before you jump in and start buying the latest and greatest machine learning application is to first look at all your data and start assessing uh how reliable it is how it is stored is it all accessible are there data sources that perhaps you can combine together and first start looking for some quick wins to see what it is that you can do with this information and the final thought i want to leave the thought i want to leave you with is the issue of data literacy so with all this technology and all these developments in computing power and open source software and and tools to analyze data i believe that this requires a higher level of data literacy than we expect than we would have previously previously expected from our staff so the digital revolution requires people to increase their knowledge of how to use a computer and how to use data and i believe that it skills are no longer the domain of the it department now normally when people say oh computer programming they immediately turn to the it department and believe that that's where it has to be solved and i strongly advocate the fact that border engineers and other water professionals should learn computer science and in order to become data scientists and i think water utilities are in a great position to really embrace the benefits of data science because we have so many great engineers in our industry and if you teach them some computer science they will be able to do great things with those tools that are available to them and to promote this idea i developed a course called data science for water professionals and anyone that's listening to this talk anyone who signed up for the for the conference can get free access to this course and also this course will teach you the principles of data science the principles of good data visualization etc and how did how to do dialysis strategically and it teaches you the basic principles of how to solve certain problems with the r language of statistical computing so if you're interested go to the website and register for free i like to leave you with a final wisdom one of the largest buildings in bendigo is a buddhist stupa so that's a large buddhist temple it's quite quaint to see this temple in the middle of the australian bush surrounded by hopping kangaroos but the great stupid of universal universal compassion is a great building here in regional victoria and reading about buddhism i found this amazing phrase that i think applies to digital transformation award utilities or any other business for the matter and that is the buddha said that to know and not to do is not yet to know in other words collecting a lot of data and analyzing data and visualizing it is great it's great to know things but really the key to a digital transformation is doing something with that information and because wall detail is always a physical business the end result will always be that you either you have to go out and you have to do physical things to improve your network this is the key of digital transformation i thank you for your attention and listening to my talk if you like to connect with me you can find me on my website lucidmanager.org or on twitter which is also a lucid manager or perhaps you can connect with me on linkedin and i really hope to be able to meet some of you because i was really looking forward to being in johannesburg the place where i started my career but it's another story for another day um if you're interested to talk with me further about digital transformation you're more welcome to contact me and thanks for your time

Channel: The Lucid Manager

Views: 19

Rating: 0 out of 5

Keywords:

Id: uLmy0aMXfPQ

Length: 25min 31sec (1531 seconds)

Published: Mon Nov 02 2020