Indonesia Conference Directory

Analog meter recording devices that have worked in the industry need to be optimized to be able to work in an integrated manner in the era of the industrial revolution 4.0 in Indonesia. An intelligent device that can read analog data and convert it to digital data needs to be made to make system integration in the industry easier. In this paper, we propose an IoT system that is built using Raspberry, a single board computer, and a camera sensor. The camera used is a low resolution USB camera with a size of 640 * 480 pixels. Retrieval of analog meter image data is done in a closed environment with a light source from the LED. The analog meter object used is a square meter analog object with a red pointer. Analog meter scale values in position angles between 45 degrees to 135 degrees. The transformation of analog data values into digital data was successfully carried out by the red masking process, followed by the edge detection process with the Canny method, and continued with the line detection process using the Hough Transform method. This digital data was also successfully transmitted via the Raspberry WIFI network to Firebase cloud database storage. From the experimental results, the maximum absolute error from the device is 0.333 or 2 degrees different from the analog meter reading angle. From all experimental data, the standard error of device is 0.0347. With these results it can be concluded that an IoT device can be made using a Raspberry and a camera sensor to read analog meters with a fairly high degree of accuracy.