INTERNET OF THINGS (IoT) BASED SCALES FOR DYNAMIC LOADS WITH WEIGHT FORECASTING FEATURE

Abstract

A scale is a measuring instrument used to measure the weight of an object. The current weighing system has not been able to identify the item being weighed and has not been able to save the weighting data automatically, in this study a weighing system was developed that can identify the load using RFID and can store the weighing data in a web browser so that there is no manipulation of the weighing data. and the system is able to predict the results of weighing the load with a linear regression method to determine the growth of the weight object. The microcontroller used is Arduino Uno and the weight sensor used is a half bridge type load cell. From the sensor accuracy test with a load weight of 0.90 Kg, 12.95 Kg, 55.42 Kg, 125.95 Kg, the highest error occurred when the load was 0.90 Kg while the lowest error was at 55.45 Kg. the object of a moving load is difficult to produce the same weight in the first measurement and so on. Testing the accuracy of predictions is evidenced by the MSE value of 0.0017 and MAPE of 0.0742%, which means that the prediction using the linear regression method is classified as very good.