Design and Implementation of SLAM Algorithm on Miniature Waiter Robot for Autonomous Navigation in LiDAR Based Indoor Environment

Fikri Aditya Maajid

Abstrak

This study presents the design, development, and experimental implementation of an autonomous miniature servant robot equipped with LiDAR and multiple sensors for indoor navigation using the Simultaneous Localization and Mapping (SLAM) algorithm. With the increasing demand for automation in service sectors such as restaurants and cafes, the study explores affordable yet reliable solutions by integrating robotics, real-time mapping, and autonomous control. The robotic system is built on top of the Raspberry Pi 5 as the main processing unit, connected to a series of sensors including a TFmini LiDAR sensor for distance mapping, a BNO055 IMU for orientation estimation, and a wheel encoder for odometry. The robot is designed to navigate and operate in a structured indoor test arena that simulates a small-scale restaurant environment. Two SLAM algorithms were implemented and compared in this study: ORB-SLAM, a feature-based visual SLAM method, and Google Cartographer, a LiDAR-based SLAM method. Both algorithms are evaluated in terms of mapping accuracy, computational efficiency, and real-time performance on embedded systems. The system operates under the ROS 2 framework and features a full-stack control interface via a Flask API and a web-based GUI using Node.js, enabling real-time destination commands, path visualization, and route execution back using custom JSON-based navigation data. Experimental results showed that Cartographer performed better in low-light LiDAR dominant scenarios with a mapping error rate of below 4.2%, while ORBSLAM provided richer environmental features but was limited by Raspberry Pi computational constraints. The integration of IMU and encoder data significantly improves the estimated pose during dynamic movement. In addition, the specially designed robot successfully executes predefined delivery and return paths, with real-time feedback available through an easy-to-use web dashboard. The project contributed to the development of affordable low-power autonomous service robots for indoor applications. It lays a practical foundation for future research in hybrid SLAM, AI-assisted pathway planning, and multi-robot coordination in dynamic environments.

Kata kunci

© 2026 UPA Perpustakaan POLINEMAStatus LayananKebijakan PrivasiSyarat & Ketentuan

v2026.11.17.1 · 828354d