Development of Component Recognition Applications and Labor Tools Based on Android and Tiny Yolo Network (Case Study: Signal and System Laboratory)

Practicum activities in the laboratory usually equipped with tools and components that must be prepared in advance. This study aims to develop an application for recognizing laboratory tools and components. The application is designed for Android-baced devices by utilizing the smartphone camera and developed using Tiny YOLO. The development follows System Development Life Cycle (SDLC) methodology using waterfall model. The system then tested by training data on 1,666 image objects obtained from Google in the form of laboratory tools and components such as Arduino, Raspberry Pi, HC-05 sensor, Esp-32 Module, Multimeter, Oscilloscope, and Function Generator. The results showed that the system can detect components and laboratory tools at an optimal distance of 25-35 cm and the accuracy of object detection is influenced by the light conditions in the. From several components tested, the object detection accuracy rate for Arduino Uno is 73.33%, Raspberry Pi is 82.5%, Bluetooth HC-05 module is 86.84%, Esp32 module is 84.37%, Multimeter is 80.6%, Oscilloscope is 76.31% and 80% function generator.