Oil Well Monitoring System Based on IoT Technology and Machine Learning

The process of crude oil mining in oil wells takes a long time and requires good supervision to avoid unwanted things. This process requires full 24-hours monitoring of oil parameters such as oil temperature and flow rate. Currently, the supervision process is still done manually which may occur some errors. Based on this fact, this paper aims to design a surveillance or monitoring system that is more effective and efficient. The testing of this monitoring system uses an oil pump machine prototype with the assistance of a MAX-6675 temperature sensor and an ultrasonic flowmeter TUF-2000m as well as a sensor TM-1 transducer as an input tool. Raspberry Pi 3 as a microcontroller and a web application as an output that displays data in the form of graphs. The test stage is carried out by heating the temperature sensor, slowing down the flow of oil on the prototype, and checking the values displayed on the graph. The results of the test when the temperature sensor received heat, the microcontroller ran well, as evidenced by the data that was successfully stored on the web server and a graph showing the increase in the oil temperature value. Likewise with the flow rate sensor when it receives resistance or the flow is slowed down, the graph shows a decrease in the flow rate value in the oil. With the results of this test, the prototype of the monitoring system on oil wells with the Internet of Things (IoT) technology runs as expected, namely being able to monitor the value of oil parameters in real-time which allows effectiveness and efficiency in work to increase.