{"id":569340,"date":"2025-11-21T19:22:00","date_gmt":"2025-11-21T22:22:00","guid":{"rendered":"https:\/\/revistapesquisa.fapesp.br\/?p=569340"},"modified":"2025-11-21T19:22:00","modified_gmt":"2025-11-21T22:22:00","slug":"new-technologies-to-reduce-water-waste","status":"publish","type":"post","link":"https:\/\/revistapesquisa.fapesp.br\/en\/new-technologies-to-reduce-water-waste\/","title":{"rendered":"New technologies to reduce water waste"},"content":{"rendered":"

Sensors, fiber optics, satellite imaging, robots, and AI are all part of a growing arsenal of tools being deployed to track down hidden leaks in Brazil\u2019s water mains. Historical waste figures make the urgency clear: a June 2024 report from the Trata Brasil Institute found that the water lost in distribution systems each year could supply 54 million people for a full year. Every day, that\u2019s enough wasted water to fill more than 7,600 Olympic swimming pools, equivalent to 37.8% of all treated supply, according to Brazil\u2019s National Sanitation Information System. Meanwhile, 32 million people still lack reliable access to clean water.<\/p>\n

There is a silver lining: after six straight years of rising losses, 2024 saw the first decline\u2014down from 40.3% in 2021. But the nation still lags far behind the federal target of keeping losses under 25% by 2033. \u201cIn most industries, a 25% loss rate would be unthinkable. Yet even that goal will be tough to reach,\u201d says Maria Mercedes Gamboa Medina, a civil engineering professor at the University of S\u00e3o Paulo at S\u00e3o Carlos School of Engineering (EESC-USP). Leakage rates exceed 40% across much of northern and northeastern Brazil. In the state of Amap\u00e1, the figure is as high as 71%.<\/p>\n

\u201cThe new Sanitation Framework has turned up the pressure to get leaks under control,\u201d notes Fabr\u00edcio C\u00e9sar Lobato de Almeida, a professor of mechanical engineering at S\u00e3o Paulo State University (UNESP) in Bauru. The 2020 sanitation law (No. 14,026) outlines goals on achieving economic sustainability in service provision and on the use of technology to reduce operating costs and improve efficiency.<\/p>\n

One initiative to achieve these goals is a Surface Leak Locator (LocVas) project that Almeida is leading in partnership with the S\u00e3o Paulo state water utility, Sabesp, with funding from FAPESP\u2019s Research Partnership for Technological Innovation (PITE). The system identifies underground leaks by detecting shifts in the vibroacoustic signature of buried pipes, measured from the soil directly above them. When water escapes, it causes vibrations within a distinctive frequency range that depends on soil type, pipe material, and dimensions. Sensors placed at the surface pick up these signals\u2014a noninvasive approach that avoids the need for excavation. The team plans to test the method on a variety of pipes and ground coverings\u2014from grass to traditional Portuguese stone paving to asphalt.<\/p>\n

LocVas builds on an established leak-detection approach based on cross-correlation, which also relies on capturing the vibroacoustic signal produced by escaping water. \u201cIn a leak-noise correlator, we measure signals captured by sensors placed at two points along the pipeline. By knowing how fast the sound of a leak travels and measuring the delay between those signals, we can estimate the leak\u2019s position,\u201d Almeida explains. The difference is that traditional noise correlators must make direct contact with the water main\u2014usually through a manhole or by digging small access pits in the soil. The LocVas system, in contrast, is not required to be in contact with the pipe and can capture vibration signals at a distance. \u201cA novel feature of our method is that it also estimates the routing of the pipe itself along with the leak\u2019s location\u2014something commercial devices don\u2019t yet do,\u201d says Almeida. The team published details of the technique in 2024 in the Journal of Physics:<\/em> Conference Series<\/em>.<\/p>\n

\"\"

Stattus4<\/span>An engineer uses a 4Fluid M\u00f3vel smart listening rod to check for leaks in a municipal water lineStattus4<\/span><\/p><\/div>\n

Launched in 2022 and expected to conclude in 2026, the UNESP-led project spans the full academic ladder\u2014from undergraduates in scientific initiation to postdoctoral researchers\u2014and has already led to several spin-off studies. One example is a master\u2019s thesis by Bruno Cavenaghi at the Bauru School of Engineering (FEB), describing a project using cameras and computer vision to see the subtle vibrations of buried pipes through the ground above them. \u201cWe set the camera just inches from the ground, angled 10 to 15 degrees, to capture minute movements at that spot. Every pixel in the footage works like a tiny, virtual sensor,\u201d explains Cavenaghi. In 2024, Cavenaghi earned the Young Professional Award from the Sabesp Engineers Association (AESabesp) for this work, which he showcased at the 35th<\/sup> National Conference on Sanitation and the Environment.<\/p>\n

To test his approach, he used a leak-noise simulator developed at the university. \u201cIt\u2019s the first in the world,\u201d says Almeida, who supervised Cavenaghi. The bench simulator, Almeida explains, isn\u2019t just a research tool\u2014it is also a training platform for utility crews.<\/p>\n

An array of leak-detection solutions<\/strong>
\nAcoustic detection is still the go-to method for finding water leaks. The most basic\u2014and oldest\u2014tools are listening rods and geophones. A listening rod, essentially a metal probe, is often used for an initial sweep: the operator presses it against hydrants, service connections (the junction between the mains and a building\u2019s plumbing), or other access points to capture vibrations caused by a leak. For pinpoint accuracy, crews use geophones, devices that work much like a doctor\u2019s stethoscope. Geophones amplify underground sounds but require a skilled ear to interpret the captured noise.<\/p>\n

To speed up this process\u2014and make it practicable even in regions short on trained personnel\u2014Brazilian startup Stattus4 designed an intelligent, AI-driven detection system. In 2018, three years after its founding, the Sorocaba-based company launched 4Fluid M\u00f3vel with funding from the FAPESP Innovative Research in Small Businesses (PIPE) program. The device can be operated by workers with minimal field experience.<\/p>\n

Noise captured through a listening rod is transmitted to the cloud, where artificial intelligence [AI] software compares it against a database of previously cataloged leak sounds. \u201cWe\u2019ve logged over 7 million acoustic samples, which help our AI software detect potential leaks with better than 80% accuracy,\u201d says cofounder and CEO Mar\u00edlia Lara. Some of Stattus4\u2019s clients include major utilities such as Sabesp; Copasa in Minas Gerais; Sanepar in Paran\u00e1; and the \u00c1guas do Brasil Group, which serves 32 municipalities across S\u00e3o Paulo, Minas Gerais, and Rio de Janeiro.<\/p>\n

Founded just four years ago in S\u00e3o Paulo, Waterlog is also leveraging AI to detect the sound signatures of leaks. Chemical engineer and cofounder Fernando Loureiro Pecoraro says Brazil\u2019s amended sanitation framework led to a collaboration with utility companies to develop a high-tech anti-leak system, called Iris. \u201cIts key advantage is real-time monitoring,\u201d Pecoraro says. \u201cMounted at the service connection, the system listens for noise in the network and flags a leak the moment it happens. To pinpoint the exact location, you still need other tools, like geophones.\u201d<\/p>\n

While acoustic methods dominate, the leak-detection market offers a growing menu of technological options. Choosing the right one often comes down to cost-effectiveness\u2014a calculation that varies with local economic conditions and available technology. According to C\u00edcero Mirab\u00f4 Rocha, a civil engineer in Sabesp\u2019s Operational Development division, the utility routinely evaluates pitches from tech companies. \u201cWe always start with a trial,\u201d he says. \u201cIf it works, it goes into our toolbox. But there\u2019s no silver bullet\u2014the only thing that works is a combination of techniques.\u201d<\/p>\n<\/div>

\n \n \n \n <\/picture>Alexandre Affonso\/Revista Pesquisa FAPESP<\/span><\/div>
\n

In April, Sabesp began piloting a leak-detection system from Israeli firm Asterra that uses satellite imaging and remote sensing. The technology beams a wave signal into the ground\u2014up to 3 meters (m) deep\u2014to detect traces of chlorine, a chemical added during water treatment. When paired with AI analysis, the data is turned into maps highlighting areas likely to be saturated with leaking water. In a 50-kilometer stretch of water mains in the S\u00e3o Paulo metro area, the satellite-based method spotted 81 leaks\u2014compared to just 14 found using conventional tools. \u201cPinpointing the exact spot still requires acoustic methods,\u201d Rocha notes.<\/p>\n

For inspecting the insides of large-diameter mains\u2014the pipelines that move treated water from plants to distribution reservoirs\u2014utilities are also turning to robotics. One provider is Yadah Robotics, based in S\u00e3o Jos\u00e9 dos Campos in S\u00e3o Paulo State. Founded in 2015, the company has built four types of robots capable of video-inspecting sewer and stormwater pipes ranging from 15 centimeters to 2 m in diameter. \u201cWe\u2019ve served city governments, water utilities, and industrial clients,\u201d says mechanical engineer and CEO Fernando Sato.<\/p>\n

From bench to business<\/strong>
\nYadah Robotics\u2019s homegrown machines are a rare exception in Brazil\u2019s leak-prevention landscape, where most devices are designed overseas. \u201cWe still rely heavily on imports\u2014we might use a Japanese geophone or a British correlator,\u201d says civil engineer Marcelo Kenji Miki of Sabesp\u2019s Sewage Treatment Division.<\/p>\n

From 2015 to 2019, while managing Sabesp\u2019s Research, Development, and Innovation Projects Department, Miki helped develop Latin America\u2019s first domestically built leak-noise correlator in a collaboration with UNESP, led by mechanical engineer Michael John Brennan. \u201cWe connected the field crews with the research labs so researchers and operators could learn from each other,\u201d Miki recalls.<\/p>\n

\"\"

Davi Silvano Costa da Rocha\u2009\/\u2009Yadah Robotics<\/span>A Yadah Robotics unit performing a video inspection inside a 15-centimeter PVC pipeDavi Silvano Costa da Rocha\u2009\/\u2009Yadah Robotics<\/span><\/p><\/div>\n

The result was a patented device tailored to Brazil\u2019s soil conditions, built at roughly one-tenth the cost of its imported counterpart. More recently, the team earned an international patent for an innovative signal-processing method that calculates the time delay between captured sounds\u2014a key factor in pinpointing leaks with high precision.<\/p>\n

\u201cThe one thing we couldn\u2019t do,\u201d Miki admits, \u201cwas turn our research into a market-ready product.\u201d Almeida of UNESP still hopes Brazil\u2019s homegrown leak-noise correlator will make it to market. \u201cWe\u2019ve restarted the project and have already spoken with a potential industry partner,\u201d he says.<\/p>\n

The above interview was published with the title “Tracking down water waste<\/strong>” in issue 353 of July\/2025.<\/p>\n

Projects
\n1.<\/strong> LOCVAS \u2013 Surface Leak Locator (n\u00b0 20\/12251-1<\/a>); Grant Mechanism<\/strong> Research Partnership for Technological Innovation (PITE); Cooperation agreement<\/strong> SABESP; Principal Investigator<\/strong> Fabr\u00edcio C\u00e9sar Lobato de Almeida (UNESP); Investment<\/strong> R$1,189,856.89.
\n2.<\/strong> Investigation of ground vibration measurements for leak detection in buried water pipes using cameras (n\u00b0 22\/07586-0<\/a>); Grant Mechanism<\/strong> Master’s Fellowship; Cooperation agreement<\/strong> SABESP; Supervisor<\/strong> Fabr\u00edcio C\u00e9sar Lobato de Almeida (UNESP); Beneficiary<\/strong> Bruno Cavenaghi Campos; Investment<\/strong> R$135,231.55.
\n3.<\/strong> Automatic leak detection system for the water distribution network (n\u00b0 17\/00798-3<\/a>); Grant Mechanism<\/strong> Innovative Research in Small Businesses (PIPE); Principal Investigator<\/strong> Antonio Carlos Oliveira J\u00fanior (Stattus4); Investment<\/strong> R$239,236.29.
\n4.<\/strong> Development of a Brazilian signal correlator optimized for locating and detecting leaks in SABESP’s underground water pipes, in conjunction with effective tools for training leak detection teams (n\u00b0 13\/50412-3<\/a>); Grant Mechanism<\/strong> Research Partnership for Technological Innovation (PITE); Cooperation agreement<\/strong> SABESP; Principal Investigator<\/strong> Michael John Brennan (UNESP); Investment<\/strong> R$931,898.66.<\/p>\n

Scientific articles<\/strong>
\nMATOS, P. H. et al.<\/em> Enhancing leak location in buried water pipes using array signal processing techniques: The effect of wave velocity variation<\/a>. Journal of Physics: Conference Series<\/strong>. Vol. 2647, no. 8. 2024.
\nISLAM, M. R. et al.<\/em> A review on current technologies and future direction of water leakage detection<\/a>. IEEE Access<\/strong>. Vol. 10. Oct. 2022.
\nCAMPOS, B. C. Uso de c\u00e2meras e t\u00e9cnicas de vis\u00e3o computacional aliadas a m\u00e9todos vibroac\u00fasticos na localiza\u00e7\u00e3o de vazamento em tubos enterrados<\/a>. Anais do Encontro T\u00e9cnico AESabesp e do 35\u00ba Congresso Nacional de Saneamento e Meio Ambiente<\/strong>. 2024.<\/p>\n","protected":false},"excerpt":{"rendered":"Universities and companies invest in solutions for detecting leaks in water distribution networks","protected":false},"author":131,"featured_media":569376,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_exactmetrics_skip_tracking":false,"_exactmetrics_sitenote_active":false,"_exactmetrics_sitenote_note":"","_exactmetrics_sitenote_category":0,"footnotes":""},"categories":[169],"tags":[228],"coauthors":[440],"class_list":["post-569340","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-technology","tag-engineering"],"acf":[],"_links":{"self":[{"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/posts\/569340","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/users\/131"}],"replies":[{"embeddable":true,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/comments?post=569340"}],"version-history":[{"count":3,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/posts\/569340\/revisions"}],"predecessor-version":[{"id":570126,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/posts\/569340\/revisions\/570126"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/media\/569376"}],"wp:attachment":[{"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/media?parent=569340"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/categories?post=569340"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/tags?post=569340"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/revistapesquisa.fapesp.br\/en\/wp-json\/wp\/v2\/coauthors?post=569340"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}