AI for Water Supply Systems: Planning for the Future
Case Study: MPWiK Wrocław x Deloitte – Predictive Maintenance
About the Project:
Can failures be predicted before they occur to avoid costly repairs? Yes! A modern predictive system for managing the water supply and sewage network has been implemented in Wrocław to efficiently detect risks of malfunction and prevent problems.
In this case study, we will look at how technology has changed the way water and sewage infrastructure is managed, improving not only operational efficiency but also the quality of life for city residents.
Modern cities face challenges related to maintaining water supply and sewage infrastructure, and failures can lead to high repair costs, water losses, and dissatisfaction of residents. In response to these problems, MPWiK Wrocław (the municipal water supply and sewerage company), in collaboration with Deloitte and AWS (Amazon Web Services), decided to implement a Predictive Maintenance system. The system utilizes advanced data analysis and artificial intelligence algorithms.
“MPWiK Wrocław manages a water supply network of over 2,000 km and thorough rehabilitation of such extensive infrastructure would require almost unlimited financial resources. Given budget constraints, it is important to prioritize the sections that require rehabilitation. In view of these challenges, it became necessary to find a tool allowing effective planning for modernization and achieving the best results with available resources”.
– Piotr Słomianny, CIO & CFO, MPWiK S.A. in Wrocław
Get inspired! Watch our video success story
RISK CAUSE ANALYSIS
A tool developed by Deloitte and MPWiK S.A. in cooperation with the technology partner – Amazon Web Services combines advanced data analytics with deep understanding of the specifics of critical infrastructure. Key steps in the project implementation included:
- integrating data from existing internal data sources (e.g. pipe type, age, material) with new information sources (e.g. location of other facilities);
- creating an integrated database containing over 300 variables for each infrastructure element;
- developing and testing many machine-learning models to identify non-obvious data patterns, e.g. how proximity to roads or tracks impacts the risk of failure.
“The system developed in cooperation with Deloitte, MPWiK S.A. and AWS, enables intelligent planning of investments and modernization. Based on its predictive module, it analyzes the probability of failures in various parts of the network and estimates the required repair budgets. It utilizes advanced AWS cloud technologies which form the foundation of its operations. With the flexibility and scalability of the cloud, it can process vast amounts of data from water and sewage networks regardless of their complexity and location.”
– Dawid Dybuk, Key Account Manager, Amazon Web Services
“Our solution uses advanced AI algorithms to analyze hundreds of thousands of data points, allowing us to predict potential failures with up to 90% accuracy. The system takes into account variables such as the age of the transmission pipe, the material it is made from, and hundreds of variables that describe the environment in which the section is located, such as those causing higher probability of network failure or the proximity of tram tracks which also negatively affects the network. As a result, MPWiK Wrocław has a tool that makes it possible not only to respond to failures but also to avoid them”.
– Amadeusz Andrzejewski, AI&Data Leader, Deloitte.
PRECISE PREDICTION OF FAILURES
One of the key functionalities of the tool is its interactive map which presents the risk of damage. The map visualizes all elements of the infrastructure and marks them with colours that reflect the associated level of failure risk. The user can select any element to access detailed information, such as estimated repair costs, potential failure costs or forecasted risk for different periods. The map also offers filtering capabilities by selected variables and data export to formats compatible with other analytical tools. Along with the tool, users are given the ability to update data — for example, in connection with changes in material prices, the construction of a new section of water supply system, or the modernization of infrastructure.
The tool's capabilities do not stop at responding to ‘what can break?’. They make it possible to answer the question ‘why and what can we do about it?’ thus opening up new opportunities in infrastructure management. Furthermore, the tool enables users to export data independently.
A key functionality of Predictive Maintenance is its comprehensive financial model, allowing transformation of technical data into specific business decisions. The model analyzes two main scenarios: planned modernization and unplanned failure repair, considering over 50 expenditure components, such as the costs of labour, materials (e.g. pipes or restored road surfaces), and land occupation. Analyzing and comparing different scenarios allows for budget optimization and prioritizing investments, maximizing the efficiency of all invested resources.
DEVELOPMENT AND SCALABILITY
Based on the obtained results, experts plan further development of the Predictive Maintenance tool to increase its functionality and availability. The planned improvements include adding different language versions to the system. An update of the financial module for more accurate analysis of operational costs and potential savings is also being considered. Additionally, the UX layer is to be enhanced with interactive charts and visualizations which will facilitate data interpretation, expedite decision-making, and improve infrastructure management.
Further development of the tool opens up new possibilities in the field of predictive maintenance of urban infrastructure and transitioning from a response-based management model to a proactive one, contributing to even greater operational efficiency and sustainable resource management.
“We supported the implementation of every stage of the project – from requirement analysis to team training. Our solution is scalable. We can adjust it to different network sizes which means that the tool is effective for both larger cities and smaller municipalities. In the future, we plan to add new functionalities, such as implementing a predictive module for the heating or gas network and integration with an ERP system and budget module.”
– Łukasz Kowalski, Business Development Director, Deloitte
About MPWiK WROCŁAW
MPWiK S.A. in Wrocław has a history spanning nearly 150 years. Today, the company supplies high-quality water from the Oława and Nysa Kłodzka rivers to around one million residents in Wrocław and seven neighboring municipalities.
For years, MPWiK S.A. has been committed to protecting the natural environment and engaging in social responsibility, which sets it apart from other economic organizations. The company also takes actions aimed at raising environmental awareness—for example, through the #DrinkTapWater social campaign. A particularly noteworthy example of its CSR policy in action is the promotion of ecological attitudes among the youngest generations through the creation of the H2O Academy. This initiative is aimed at second-grade primary school students and includes all primary schools in Wrocław. Over the three years of the project, around 20,000 children in Wrocław have taken part. In addition, residents can find a wealth of information relevant to the local community on the website www.mpwik.wroc.pl as well as on a dedicated platform designed specifically for Wrocław’s inhabitants. www.pijkranówkę.pl
About Amazon Web Services
Since 2006, Amazon Web Services (AWS) has been the world’s most comprehensive and widely adopted cloud platform. AWS continuously expands its services to support virtually any cloud workload, currently offering over 240 fully featured services for computing, storage, databases, networking, analytics, machine learning and artificial intelligence (AI), Internet of Things (IoT), mobile, security, hybrid solutions, and application development, deployment, and management—across 114 Availability Zones (AZs) in 36 geographic regions, with plans announced for 12 more Availability Zones and four new AWS Regions in New Zealand, the Kingdom of Saudi Arabia, Taiwan, and the AWS European Sovereign Cloud region. Millions of customers—including the fastest-growing startups, largest enterprises, and leading government agencies—trust AWS to power their infrastructure, increase agility, and reduce costs. For more information about AWS, visit aws.amazon.com
Opens in new window
