Artificial intelligence is a new alias for combating water vapors

PORTO ALEGRE, RS (FOLHAPRESS) – According to the National Health Information System (SNIS), 40% of water destined for consumption is lost in distribution. Seriously or enough to supply an estimated 66 million people in a year, nearly double of Brazil’s 35 million lack access to clean water. An artificial intelligence can make it cenário.

The work to meet vazamentos is slow and inefficient. As supply companies we use a geophone, an electrical device with an acoustic system, which works as a sort of “asphalt stethoscope”, capable of “shielding” or damaging two joints and always detecting sudden changes.

Looking for a more efficient technology, Marília Lara and Antônio Oliveira created a startup Stattus4, in Sorocaba (interior de SP). The first device developed was an artificial intelligence system called 4Fluid Móvel, which can identify damage in underground troubles and generate area reports.

Unlike the geofone, which needs a technician to work, the device can be used for anything. It is a coletor with a hurry, which, ao ser encostada no hidrômetro, capta in dez segundos os sons nos canos.

Keeping to analyze, a supply company has to send functionality to the local problem. “The team does not specify ever more rodar a cidade in search of travels. The hardware guarantees that the nut is accurate. Our product is two times faster than a geophone and, on average, this half the price”, says Marília.

The value of the service varies according to the number of supply concessions, the number of inhabitants of the municipality and the quantity of vazzamenti in the region.

“Hoje, 53 olympic pools are saving all the dias for our technology. The goal is to chegar a mil por dia in três anos,” says Marília.

Among the clients is the Grupo Águas do Brasil, which expects cidades such as Niterói (RJ), Petrópolis (RJ) and Votorantim (SP). Somadas, as três têm close to 940 million people.

In general, as water leaks occur for a long time from rede and dos ramais de distribution, frutos de vazamentos nas tubulações or problems such as miscalibration of hydrometers, reading errors, frauds and clandestine connections.

Identify and prevent vazamentos na rede foi o desafio aceito pelo então bolsista do Laboratório de Eficiência Energética e Hidráulica em Saneamento (LENHS) da UFPB (Universidade Federal da Paraíba) Thommas Flores.

“Com a IA (artificial inteligência) dá para pegar o padrão de consuming de um sistema de abastecimento e compará-lo ao sumption de qualquer instante, e assim identify vazamentos”, diz Flores, hoje mestre in engenharia elétrica.

The project was created with an algorithm and was tested in the water reservoir of the UFPB laboratory. I planned to apply João Pessoa’s (PB) network system. Cerca de 820 mil pessoas seriam impactadas.

Leave a Reply

Your email address will not be published. Required fields are marked *