Inspiration
The inspiration for this project arose from the growing inefficiency and water losses observed in water supply networks, particularly those caused by unexpected pipeline failures and ruptures. In Portugal, a significant amount of data related to water infrastructure performance is already collected through systems such as RASARP reports; however, this information is often fragmented and analysed in isolation. This project was motivated by the need to transform historical and heterogeneous data into actionable knowledge, enabling a shift from reactive maintenance strategies to predictive and preventive decision-making. The integration of artificial intelligence and IoT technologies emerged as a natural solution to anticipate failures, reduce water losses, and improve the sustainability and resilience of water supply systems.
What it does
This project aims to develop an artificial intelligence–based system for the predictive assessment of water supply infrastructure performance in Portugal. By integrating and cross-analysing historical datasets from multi-year RASARP reports from the 21st century, the system will consolidate heterogeneous technical, operational, and performance indicators related to water pipelines and the entities responsible for their management.
The project focuses on the classification and evaluation of key variables such as material type, pressure conditions, environmental context, age of infrastructure, and historical performance according to predefined performance and risk parameters. Through advanced data analytics and machine learning techniques, the system will identify patterns associated with degradation, failure probability, and structural vulnerability.
Based on this integrated analysis, the project will generate spatial risk maps that distinguish high-risk and low-risk zones, with particular attention to critical areas such as pipelines located beneath roads, zones with high operating pressure, and sections where material characteristics increase failure susceptibility.
The ultimate objective is to support the deployment of IoT-based monitoring systems in the most critical locations, enabling early detection of anomalies and prediction of failures before they occur. By shifting from reactive to predictive maintenance, the project aims to significantly reduce water losses, improve network reliability, and support more efficient and sustainable water resource management.
How we built it
We built it using a site designed to create AI`s programs and used another tools to make the tests and check AI ´s performance.
Challenges we ran into
We had some problems on choosing what would we do , since we liked this idea but we haven`t had many contact in programming.
Accomplishments that we're proud of
We are proud of developing something on our own from the very start and something that can be useful in the real world.
What we learned
We learned programming skills, gained valuable insights into the complexity of water supply infrastructure management, and many more such as technical skills in data integration, risk classification ant machine learning applied to infrasestructure systems.
Log in or sign up for Devpost to join the conversation.