We have recently learned about some consumption data in the Region of Madrid that are quite remarkable. Water consumption in the region is 7.4% lower than 10 years ago, even though the population has increased by 7%.
My first thought when I heard this was: finally, people are realising we need to save water! Finally, people are closing the tap when they brush their teeth or shave, and are taking showers instead of baths. The truth is that the savings are mostly due to the detection of leaks in the water distribution network.
From our point of view, the most interesting thing is not the savings data per se, which is excellent news, but the fact that leaks are being prevented.
Currently, to prevent those leaks, data collected throughout the past 20 years are analysed and cross-checked to see which pipelines need replacement. That is, the current system does not detect leaks, but it replaces old pipelines with new ones in a certain order, according to statistical data.
This means that the system still offers opportunities to use new technologies, such as the Internet of Things, IoT. By IoT we mean a series of devices connected to the Internet that collect data massively and are distributed throughout the network.
We can think about a system that can monitor in real time what is happening at each point of the water network and, based on the data collected, we would know when there is a leak.
Currently, the conditions are present for this to happen:
1. IoT devices (sensors) which are small and inexpensive, with very low energy consumption.
2. Capacity to send data in real time and in large amounts.
3. Tools to store and analyse information in the cloud.
4. Professionals trained to perform the analyses.
In this scenario, we face two great challenges. On one hand, there is energy consumption. The sensors have to be placed strategically and due to the nature of the system (underground pipes), they cannot be regularly changed. We have to create a system of long-lasting sensors. In the IoT sector, there are companies that already produce measuring sensors with batteries that last up to 10 years.
The other great challenge is knowing what to measure in order to determine if there is a leak at a particular location. The combination of water flow, speed and size of pipes are some of the parameters we have to take into account to establish whether there is a leak. It is not simple at all, and it is necessary to rely on technical staff and experts in the area to determine the best way of measuring.
Sensors would collect information constantly, which would then be sent for subsequent analysis and processing. The data collected would be transmitted using wireless technology. In this case, a good means of communication is the Zigbee protocol, which allows communication in a 10-20 m range, with very low energy consumption, even lower than with Bluetooth, and at a low cost.
The next point to consider is data storage. Nowadays there are enough cloud-based storage solutions that allow collecting huge amounts of data at reasonable prices. The cost of cloud-based storage is getting lower and lower, so it is not an obstacle for the project.
Finally, a data science team must analyse the data obtained and establish some parameters to assess the status of pipelines and detect potential leaks. If afterwards we add to the system the capacity to learn automatically, we could, in a second phase, realise that a pipeline needs changing before a leak takes place.
The current way of working is effective, but is based on data from the past. A system such as the one described above would be more effective and increase water savings in the future.
Comments