Non-revenue water in LATAM: The role of AI in cutting water losses

Water leaks in Latin America are a huge problem for the region. According to the World Bank, 45% of water is lost before it reaches consumers. These figures match those published in a Development Bank of Latin America and the Caribbean (CAF) study on water safety, which showed that 26 cities sustained losses which exceeded 60%.

This data on non-revenue water (NRW) in Latin America underscores the need to roll out measures to make water resource management more efficient. This is especially true at a time when the region is experiencing prolonged droughts such as the ones in Mexico, Uruguay, and Brazil. In fact, the Peruvian National Superintendence of Sanitation Services (SUNASS) has already warned that Peru, Chile, Mexico, the Dominican Republic, and Argentina will be the countries hardest hit by water stress in just five years.

Francisco Eduardo Hernández, Idrica LATAM’s Business Development Manager, thinks that the problem of NRW in most Latin American countries “lies in aging infrastructure, failure to invest in upgrading, and inefficient management which has led to hugely inefficient distribution systems.”

He pointed out, however, that measures such as harnessing artificial intelligence and the Internet of Things “are already enabling operators to cut losses, step up efficiency, and ensure fairer access to the resource.”

Applying AI: Three use cases 

Technology is proving to be a turning point in the battle against NRW. AI-based systems, known as smart water engines, coupled with data lake platforms, enable large volumes of water data to be gathered, integrated, and analyzed in real time, improving water management. As Silvia Escamilla, Xylem LATAM’s Business Development Manager, noted, “adding AI to water utilities provides more flexible decision-making and streamlines operational, administrative, and resource management processes, enhancing distribution efficiency.”

As more operators embrace these technologies, a steady reduction in NRW is expected, bringing significant economic and environmental benefits

The following use cases illustrate how applying artificial intelligence can be extremely beneficial in managing NRW:

Detecting leaks

This is the clearest case, where deploying artificial intelligence helps to monitor the network and identify leakage points more quickly and efficiently through smart pressure and flow control. An example is the utility Servicios de Agua y Drenaje de Monterrey (SADM), which cut water losses and achieved overall water savings of 17%, rising to 37% in many sectors, as a result of implementing the Xylem Vue platform.

Optimizing predictive maintenance

Utilities can harness AI to anticipate potential failures in real time by analyzing historical patterns. Digital twins that blend machine learning and artificial intelligence can reproduce the network’s performance under any circumstances, delivering a strategic water model that can predict future malfunctions and also improve decision-making.

Controlling water use

AI can also be useful for monitoring usage, as it helps to spot anything unusual, such as internal leaks, fraud and potential meter errors.

Future prospects

Silvia Escamilla believes that the key factor in the future development of this type of technology: "is mainly the partnership between governments, tech firms, and multilateral organizations, which is essential to fast-track the digital transformation of the water sector and ensure a more efficient, sustainable supply for future generations.”

AI and IoT have huge potential for water management in LATAM. As more operators embrace these technologies, a steady reduction in NRW is expected, bringing significant economic and environmental benefits. However, it is essential to overcome a series of barriers such as underinvestment and resistance to change in some institutions so as to achieve the greatest impact.