Machine learning for precise water leaks detection

Internet of Things emerged to revolutionize the technological world and our daily lives. Based on the ability to connect devices, capable of controlling and monitoring intelligent environments in order to reduce human action, these devices, being low consumption, make systems more environmentally fr...

ver descrição completa

Detalhes bibliográficos
Autor principal: Coelho, João Miguel de Jesus Alves (author)
Formato: masterThesis
Idioma:eng
Publicado em: 2021
Assuntos:
Texto completo:http://hdl.handle.net/10071/22085
País:Portugal
Oai:oai:repositorio.iscte-iul.pt:10071/22085
Descrição
Resumo:Internet of Things emerged to revolutionize the technological world and our daily lives. Based on the ability to connect devices, capable of controlling and monitoring intelligent environments in order to reduce human action, these devices, being low consumption, make systems more environmentally friendly, reducing energy costs and footprint carbon, but never putting in jeopardy the capability. Bearing in mind the poor management and the increasing scarcity of resources, there is a greater concern to monitor water systems in order to be able to detect and locate water leaks as soon as possible so that the waste is as small as possible. This dissertation presents a proposal for a system based on a wireless sensor network, designed to monitor water distribution systems, such as irrigation systems, which with the help of an Automatic Learning algorithm allows to precisely locate the place where gave the water leak. In order to obtain a capable and low-cost system, an analysis was made of several software and hardware modules so that the system, through an Android mobile application, allows the user to view information, collected by the sensors, and, consequently, deal with active way the problem of escape. The main advantage of this system and that distinguishes it from others is the Automatic Learning algorithm, which through the information that the sensors collect, learns from the system and any variation of values, alerts the user to where the leak is located.