Sistema de auxilio no resgate em àrea afetada por deslizamento

Abstract

Nos últimos anos a sociedade presenciou diversos acidentes causados por deslizamento de terra, que tiveram como consequência vítimas fatais. Pensando nisso esse TCC tem como objetivo utilizar a tecnologia aplicada a um sistema de gerenciamento, composto idealmente por um drone e um aplicativo para movimentação e gerenciamento de alarmes, que visa auxiliar os bombeiros no resgate e identificação das possíveis vítimas e objetos. As linhas de pesquisa utilizadas para a fundamentação teórica foram voltadas para causas, consequências e exemplos de deslizamentos, inteligência artificial, drones, sensoriamento e termologia. Durante o desenvolvimento, por questões de viabilidade econômica, houve a implementação de um robô móvel com sensores de distância e temperatura. O funcionamento do protótipo consiste no drone mapeando a área, e através do treinamento de imagem reconhecendo se existem pessoas sobre o solo, logo em seguida o carrinho, por meio do aplicativo, se movimenta sobre o local alertando sobre a distância dos destroços e se ocorre alteração de temperatura no solo, em caso de soterramento das vítimas. Por fim o projeto obteve êxito no reconhecimento de imagem e obtenção de temperatura do solo, no entanto foi necessário adaptar a leitura da imagem a uma webcam devido a capacidade de processamento do computador. As considerações finais apresentam melhorias futuras para o projeto.

In recent years, society has witnessed several accidents caused by landslides, which resulted in fatalities. With that in mind, this TCC presented to the faculty of Industrial Automation aims to use the technology applied to a management system, ideally composed of a drone and an application for movement and alarm management, which aims to assist firefighters in the rescue and identification of possible victims and objects. The lines of research used for the theoretical foundation were focused on causes, consequences and examples of landslides, artificial intelligence, drones, sensing and thermology. During the development, for reasons of economic viability, there was the implementation of a mobile robot with distance and temperature sensors. The functioning of the prototype consists of the drone mapping the area, and through image training recognizing if there are people on the ground, then the cart, through the application, moves over the place warning about the distance of the wreckage and if occurs alteration of temperature in the soil, in case of burial of the victims. Finally, the project was successful in image recognition and soil temperature acquisition, however it was necessary to adapt the image reading to a webcam due to the computer's processing capacity. The final considerations present future improvements for the project.

In recent years, society has witnessed several accidents caused by landslides, which resulted in fatalities. With that in mind, this TCC presented to the faculty of Industrial Automation aims to use the technology applied to a management system, ideally composed of a drone and an application for movement and alarm management, which aims to assist firefighters in the rescue and identification of possible victims and objects. The lines of research used for the theoretical foundation were focused on causes, consequences and examples of landslides, artificial intelligence, drones, sensing and thermology. During the development, for reasons of economic viability, there was the implementation of a mobile robot with distance and temperature sensors. The functioning of the prototype consists of the drone mapping the area, and through image training recognizing if there are people on the ground, then the cart, through the application, moves over the place warning about the distance of the wreckage and if occurs alteration of temperature in the soil, in case of burial of the victims. Finally, the project was successful in image recognition and soil temperature acquisition, however it was necessary to adapt the image reading to a webcam due to the computer's processing capacity. The final considerations present future improvements for the project.