FIDO AI is taking leak detection response to a new level with the development of Leak Central. It is a microcosm of systems thinking for the water industry with revolutionary potential beyond leak management.
There’s a common acronym for the leakage management process in the UK water sector. It’s PALM, which stands for Prevent, Aware, Locate, Mend.
Most digital technology advancement to this point has focussed on the areas of alert and locate. But recently there have been great strides in solutions which prevent leaks. One of these is the use of digital twin networks. A digital twin is a virtual model of a real network which uses real-time data from sensors to optimise processes and spot anomalies.
Introducing Leak Central
At FIDO we have taken a slightly different approach which treats leaks as an asset rather than a problem. We call this concept Leak Central. With Leak Central, FIDO AI gathers all the available information about a leak to create an identifiable asset. We then apply deep-learning based on intimate knowledge of the leak workflow process to analyse this explicit data and model outcomes.
This helps leak teams really understand a network’s performance and the conditions that influence its behaviour. It means they can learn faster from the wealth of valuable data and make better decisions.
In current models of leak management, data is frequently lost. A leak exists for an instant in time and you can lose valuable knowledge when you move on.
Systems thinking for water leak management
FIDO applies a unique ID to every leak to track it from first file to final fix. In Leak Central, we incorporate additional data from multiple 3rd party systems into FIDO’s super accurate analysis. This might include flow, engineering and workflow data.
All this gives a picture of a leak’s size, location, who found it and how, the water lost, material, time of year, weather, even nearby customer interactions. It’s added to FIDO’s neural network where it can be interpreted and used.
Soon you’re generating performance reports based on actual numbers rather than interpretive numbers. You can model different outcomes and understand how they operate. You can choose the best outcome for your objectives to better manage available resource according to the conditions. You can start planning maintenance and capital investment based on fact, rather than assumption. You can really start to prevent leaks
Next generation network model
Whereas Leak Central combines data about leaks, it’s really just one subset of data for specific purpose. For a water company the possibilities are far broader. We are now developing a technique we believe is unique in the water industry and building a Network Central model.
Network Central will analyse not just explicit data like sensor files, documents and workflow processes, but also implicit data. These are the linkages human beings make between separate pieces of information.
Unlike Leak Central, where such linkages and relationships are hard-coded into the decision-making, Network Central would be vastly more complex. It would make its own dynamic interpretations of links between pieces of information in separate systems that human beings would find very difficult to do. It will think for itself. The definition of systems thinking on a network scale.
More benefits of the Leak Central approach
Of course, this is still some months away. FIDO AI is extremely clever, but it’s still got a lot of learning to do. But it’s another way of capturing and using data effectively that water companies couldn’t do before.
Meanwhile Leak Central is in the final stages of testing. We’re starting to see that the system thinking approach to water leak management has other benefits too. This includes the advantages of bringing multiple data feeds into one interface.
Having worked in tech all my career I can vouch for the fact that it’s an IT support nightmare to run multiple systems, train everyone and pay for all the licenses. Updating the system, changing it or altering rostering is always a huge challenge.
With Leak Central we are start to see that evolving for the better.