The current environmental monitoring of the city’s waters highly depends on manual field sampling, and laboratory testing. This does not give a sufficiently high spatial and temporal resolution to be able to execute measures in a modern and effective way.
Current environmental monitoring of the city’s water is largely based on manual sampling in the field and analysis in laboratories. This does not provide sufficiently high spatial and temporal resolution for modern and cost-effective actions.
One of all the ways in which citizens use the city’s water is for swimming. From a health and safety perspective, the city needs to be able to provide the general public with information on the quality of this water that is as up to date as possible. The ability to measure bathing water quality in real time and notify the general public of the results is of the utmost importance if this is to be achieved. As things stand at present, it is difficult for citizens in the city to use information on bathing water quality. This is partly because data is difficult to interpret, but also because sampling and analysis are not sufficiently frequent to be useful in practice.
Bacteria in drinking water may present a potential risk of disease, which means that regular measurement and follow-up are very important. Even if samples are taken regularly, a more efficient detection method could save a lot of money and ensure that our drinking water is of the best quality possible.
Large amounts of data can be collected without taking up human resources by producing a cloud-based sensor technology that identifies changes in the water and analyses the results automatically. Near real-time updates also provide a temporal resolution that conventional environmental monitoring has not been able to achieve to date. This creates an opportunity to allocate resources to the areas where they will be most useful and gain a better understanding of the dynamics of the city’s water.
Use of an IoT platform and algorithms with AI analysis also makes multisensor analysis possible, which means that it will gradually be possible to create a network of sensors linked to one another. This could be used to track incorrectly connected drains and identify swimming pools that are at risk of bacterial contamination. Other types of analyses could be compared with the water quality results, such as stormwater discharge, road traffic, etc. Combining different datasets would increase intelligence on how and when different sources affect the city’s water.
Creating a user-friendly presentation of the data collected could make the information accessible to citizens, as well as administrations and enterprises within the city. This will give citizens themselves an opportunity to understand the dynamics of the water as well as sources of influence, and assess the quality of the water. This could relieve the burden on civil servants, who otherwise have to answer questions during peak swimming seasons, for example.
Using this technology in the drinking water network as well will allow contamination to be detected more quickly and provide early warnings of unsafe water. Sensors are already used extensively in drinking water production, but not with the same wireless and automatic design as the solution being tested within iWater
This project has embarked on phase 2 and is partly funded by Vinnova. The IoT data pipeline platform from Ericsson was completed in 2019, along with the algorithms from KTH to be tested. The sensor for measurement in surface water was launched in Lake Mälaren, and the Linköping University sensor was also ready to test and was applied in the network at Tekniska Verken in Linköping.
The data collection phase is now beginning as the sensors have been installed, and this will be followed by verification and evaluation of the algorithms and data collection.
Ericsson Research have contributed an IoT-based data solution by the name of “iWater Data Pipe”.
This orchestrates various tasks such as data collection, analysis and visualisation in an organised manner. All of these processes are automated using Click and Run technology in order to bring about the transformation from raw data to comprehensible information without manual intervention.
KTH has been working with fundamentally new machine learning technology between networks and has developed practical machine learning algorithms. These use recurrent neural networks for the iWater IoT network. KTH has tested the algorithms using data collected previously and is currently working on incorporating the code in Eriksson’s iWater Data Pipe. KTH is also developing new algorithms of the same type for measurements in the drinking water network, together with Linköping University.
Stockholm University has contributed to the geographical planning of the sensor network, which parameters are to be measured and the evaluation of historical and real-time data, as well as changes in levels measured over time.
Linköping University has developed its sensor, known as “Electronic tongue”, for detection of bacteria. This has not been used previously in drinking water production. The sensor does not measure bacteria numbers directly, but rather indicator substances that indicate contamination. As things stand at present, the sensor is adapted for drinking water measurements.
City of Stockholm (Environmental Department), Stockholm Vatten och Avfall AB (SVOA), Ericsson, Telia, KTH Royal Institute of Technology (KTH), Stockholm University (SU) and Linköping University (LiU).
- The project is being run using funding from Vinnova.
- Project owner: the Environmental Department, City of Stockholm
Åsa Andersson – City of Stockholm
Other project team
Bin Xiao – Ericsson
Carlo Fischione and Viktoria Fodor – KTH
Gia Destouni and Zahra Kalantari – SU
Mats Eriksson – LiU
Marie J. Karlsson, Joakim Strandh and Henrik Werner – Telia
Tommy Giertz – SVOA
Good water status is an important priority area for both the City of Stockholm and the EU. Sensors or other technology can be used to measure water composition in real time to ensure that the city’s watercourses are maintaining good quality and are compliant with the EU’s Water Directive. Data collected can then be made available to the general public in real time. This will make water quality in the city easy to monitor. Data produced can also be used as a basis for simulation of the effects of construction work in the vicinity of water, for example.
Contamination in recipients and the water distribution chain could be measured regularly, thereby making it possible to identify and remedy emissions early on so that any adverse impact on water quality is minimised. If the water can be measured in a number of locations, it will be easier to identify the area in which a problem has occurred; which in turn will facilitate implementation of focused measures so that impact on drinking water and bathing water is minimised and a high level of water quality can be ensured, while also using resources efficiently.