Case Study
Water Source
Reliable water, reliable technology
As populations grow and temperatures rise around the world, the need for plenty of clean, safe drinking water becomes ever more important. It’s hard to imagine in this day and age that entire communities—even here in Australia—struggle to maintain their basic human right to pure water. But it happens, every single day.

Water Source Australia, however, has set out to change things for the better. A social enterprise initiative of the Wise Foundation, Water Source Australia was founded after Graeme Wise and a team of volunteers in East Timor witnessed first-hand the plight of those affected by insufficient and polluted water supply. Besides the obvious spread of disease and illness, communities also had to deal with a social toll, such as the ways in which compromised water supplies affected women’s education opportunities. Here in sunburnt Australia, remote and indigenous communities are particularly prone to water quality problems, bringing the issue even closer to home. So the question was asked; is there a way to utilise grey, bore and surface water at a community level in order to decentralise water treatment and put a community’s potable water requirements back into its own hands?
In short, the answer was “yes.” Water Source Australia set about developing the relationships and hardware to make small-scale local water treatment a reality. The engineering and manufacture of this ‘Point of Entry’ equipment was one thing, but because the units would be located hundreds if not thousands of kilometres apart in remote locations a whole series of operational and technical challenges became apparent.

Which is where Smart Systems and our expertise in machine learning stepped in.

Just some of the technical challenges we needed to address included:
• The reliable ongoing maintenance of remote, highly specialised equipment
• Communications between locations where infrastructure is minimal or intermittent
• The collection and sending of large amounts of specific data
• Adhering to the regulations of WHO and the Australian Drinking Water Guidelines

Smart Systems developed the software and analytics to make all of the above possible, and much more. Clearly, when dealing with something as essential as clean drinking water, there is little to no margin for error. Our systems had to be resilient, reliable and easy to engage with. Chief among our priorities was establishing a fully remote monitoring capacity to ensure pure water was being produced at all times. And in tandem with that, a remote way to control the units and collect crucial data.
Clean water from the Cloud
The core of the Smart Systems solution is the Amazon Web Services (AWS) Cloud Environment. We used it to custom-build and deploy an IOT (Internet of Things) network to address—if not surpass—Water Source Australia’s stringent requirements. We fitted sensors to every valve, pump, and circuit board in each treatment unit—over 20 per site. These constantly collect large amounts of data as to the condition and output of the unit, before sending them to a central processing location here in Melbourne, as well as being available to local users via the web. Unit operators have the ability to send instructions remotely, and even interrupt automatic processes if necessary.
Liquid intelligence
Artificial Intelligence, or Machine Learning, is a critical component of our software design. In this case, not only can the treatment units self-diagnose and clean themselves (removing the need for human intervention unless absolutely necessary), they can perform a prediction maintenance algorithm. This is an impressive-sounding way of saying that the units have the capacity to predict when part failures are likely, both individually and across regions, enabling servicing and maintenance to occur in a planned, efficient manner before any failures are likely, thus eliminating downtime. It also enables multiple servicing requirements to coincide, instead of engineers having to travel to remote locations over and over.
In essence, the units are learning from their own performance data, which can be easily monitored and controlled from thousands of kilometres away at a city location.
A consistent stream
One of the key challenges we faced in development was how to ensure a consistent flow of data from each remote location to the central control point, thousands of kilometres away. In some remote locations, communication infrastructure capable of handling the large quantities of data necessary is patchy at best. Our solution was to use 4G wireless technology—but even then the reception could be intermittent at times. This led to us developing a clever local recording unit which downloaded all data without interruption, which could then be forwarded to the control centre once proper communications were re-established. A simple but important inclusion.
Water everywhere
Like all of our software, we designed it to be used, and to be easy and enjoyable in the process. Part of our job was to provide not only the city-based experts with all the data and analytics they needed to ensure a reliable supply of clean drinking water, but also the capacity for the local people—technicians, end-users, water authorities—to see what their treatment unit was up to at a glance. To that end, we developed multi-layered software, with a consumer web app for the curious, and the ability for authorities to track data, performance and analytics.
Community resilience, social stability
Unit trials are taking place in Australia and the UK, to ensure optimum robustness across a variety of environments. Full production and deployment will commence thereafter in about 2021, bringing a long-held dream to life for the Wise Foundation.

To Smart Systems, the appeal of the project was obvious. Any chance we get to use our powers for good in a world crying out for it is a chance we’ll wholeheartedly pursue. The concept of whole communities drastically changing their fortunes by producing their own pure drinking water is a simple but powerful one, and we’re proud to play our part in it.