This project was developed as a collaboration between Aguas de Barcelona and the Barcelona Supercomputing Center, leveraging advanced data processing and AI capabilities supported by the MareNostrum supercomputer.

Aguas de Barcelona manages a large and highly interconnected water distribution network, where decisions about pipe renewal are both critical and complex. Every year, the company must plan which segments of the network should be replaced, typically on a semester basis, balancing technical risk with environmental impact, social implications, and economic cost. A failure in the wrong place can lead not only to water loss, but also to disruptions affecting hospitals, schools, vulnerable populations, traffic, and local businesses.

Before this project, renewal decisions were mainly based on tabular information, expert judgment from planners and field operators, and estimated intervention durations per segment. While effective to a degree, this approach made prioritization slow, difficult to justify, and partially dependent on subjective criteria, especially when multiple high-risk segments competed for limited resources.

The goal of the project was to create a visual decision-support tool that would help technical teams plan annual renewals more effectively. The tool was conceived as exploratory and descriptive rather than fully prescriptive: it did not automate decisions, but instead provided rankings, spatial context, and recommendations that supported expert judgment. In particular, it highlighted pipe segments with a high probability of failure and suggested nearby segments that could be addressed together to optimize interventions.

The application was used by three distinct user profiles: network intervention engineers, maintenance managers, and planning managers responsible for defining the annual renewal calendar. Each profile had different priorities, but all shared the need to understand risk and impact within the broader urban context.

From a data perspective, the tool integrated heterogeneous sources. Internal data from Aguas de Barcelona described the state and characteristics of the water network, while external layers—such as transport infrastructure, traffic, hospitals, schools, and other sensitive urban services—were sourced from the city council and open data portals. Data quality varied across sources, requiring substantial preprocessing and exploratory analysis. In collaboration with domain experts, predictive models were developed to estimate the probability of pipe replacement based on technical, environmental, social, and economic criteria. This probability index became the backbone for prioritization and visual exploration.

The main design challenge was managing complexity without overwhelming users. The interface needed to combine many spatial layers operating at different scales, while avoiding visual conflicts, excessive overlap, and cognitive overload—especially when zooming and navigating across the city. To address this, the design relied on progressive disclosure, clear visual hierarchies, filtering and ranking mechanisms, and coordinated views that reduced dependence on a single overloaded map. Interaction was carefully designed to let users explore complexity step by step, rather than all at once.

The tool was developed through an iterative, user-centered process, including co-creation sessions and usability testing with different user profiles. Feedback from real planning workflows directly informed design refinements, ensuring that the system supported how decisions were actually made in practice.

The application has been used in production by Aguas de Barcelona’s technical staff. While it did not replace formal investment approval processes—which were already in place—it significantly improved planning by enabling faster decisions, reducing arbitrariness, and providing clearer visual justification when presenting plans to management. There were even discussions about creating a future public-facing version to increase transparency for citizens and municipalities, helping explain why certain areas are affected by water works and reducing friction related to traffic changes and commercial disruption.

As one technical planner summarized it:
“The tool helps me plan renewal work by highlighting the pipes most at risk and those with the highest impact if they are not replaced in time.”