Prototipe Robot Pengangkut Hasil Panen dengan Navigasi Otomatis menggunakan Line Follower dan Algoritma Dijkstra pada Simulasi Jalur Tanaman Alpukat

Abstract

This research is motivated by the declining number of agricultural workers, particularly during harvest seasons, as well as the high workload faced by farmers due to the continued use of manual methods in the transportation of harvested crops. To address this issue, a series of steps were conducted, including literature review, system design, hardware and software implementation, and system testing. This study aims to develop a prototype of a harvest-transporting robot for a simulated avocado plantation path using an automatic navigation system based on a line follower and Dijkstra's algorithm. The developed robot utilizes an ESP32 as the main microcontroller, equipped with a line sensor for navigation, a load cell for detecting the weight of the load, an HX711 module, DC motors, and an LCD to display real-time information. The testing results show that the robot was able to follow the designated path effectively, with the line follower sensor achieving an 86.67% success rate across 15 trials at three destination points. The load cell sensor recorded an average error of only 0.8% across five tests, indicating high accuracy in weight measurement. In transportation trials on three different routes, the system achieved an average time efficiency of 93.36% compared to the fastest time on each route. Furthermore, the comparison between manual algorithm calculations and program results showed consistency, indicating that the implementation of Dijkstra's algorithm worked as designed. This development is expected to support the broader adoption of automation technologies in Indonesia's agricultural sector.