Desenvolvimento e viabilidade de dispositivo para monitoramento de proximidade de objetos em operações agrícolas mecanizadas

• Ano de defesa: 2023
• Autor(a) principal: Rivas, Alex Leonel Cañar
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Sensor discreto; Máquinas agrícolas; Ultrassônico; ESP-NOW; Discrete sensor; Agricultural machinery; Ultrasonic; CNPQ::CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA
• Link de acesso: http://repositorio.ufc.br/handle/riufc/75036

In modern agriculture, agricultural machinery and equipment are entering the era of digital agriculture, requiring a high degree of expertise in the selection, operation and maintenance of agricultural machinery, technology that directly or partly affects the increase in the accident rate. With the availability of high-tech machinery and equipment, the risk of endangering the lives of human beings involved in agricultural activities is increasing. The aim of this research is to develop a device for monitoring the proximity of objects in mechanized agricultural operations and to contribute to the Laboratory for Research into Accidents involving Agricultural Machinery (LIMA) and to small farmers in preventing accidents at work with agricultural equipment. This device is based on emitting signals by measuring the distance of obstacles from the sensor. To do this, an electrical circuit was assembled using the ESP32 Wroom and programmed in the Arduino IDE development environment. The case was then designed in Fusion 360 and printed on a 3D printer. In order to assess the device's accuracy and sensitivity in terms of distance, bench tests were carried out using a 5 m tape measure, and to define the angle and distance of detection, a 360° grader was used. Next, the device was evaluated in the field with 10 cardboard obstacles to be detected at three speeds and three distances, and tests were also carried out to determine the effect of temperature and wind speed on detections. Finally, a 3k factorial design with three replications was used for statistical analysis. In the study it was determined that the sensor on the bench is capable of detecting from 50 cm to 400 cm with an overall accuracy of 0.63 standard deviation, with an acceptable sensitivity by means of a statistical control. Statistically, the device at speeds of 3 km h-1; 5 km h-1; 7 km h-1 and object detection distances of 1.5 m; 2.5 m; 3.5 m showed a difference in distance (p=0.000) and a detection limit of 2.10 m in the Pareto diagram. As for the effect of temperature and wind speed, they had no significant effect. In this way, the prototype is functional and accurate in issuing alerts so that the operator can make the best decisions to avoid injuries or accidents in agricultural tasks. In addition, the prototype can be adjusted to obtain real-time information and be incorporated into the Internet of Things (IoT).