IAES International Journal of Robotics and Automation (IJRA) Vol. 13, No. 4, December 2024, pp. 391~400 ISSN: 2722-2586, DOI: 10.11591/ijra.v13i4.pp391-400  391 Journal homepage: http://ijra.iaescore.com Experimental results on position and path control of an automated guided vehicle using fixed camera at ceiling and color markers Ramesh Pungle 1 , Atul Andhare 2 , Durva Pungle 3 1 Department of Mechanical and Automation Engineering, P.E.S. College of Engineering, Chhatrapati Sambhajinagr, India 2 Department of Mechanical Engineering, Visvesvaraya National Institute of Technology, Nagpur, India 3 Department of Computer Engineering, SCTR’s Pune Institute of Computer Technology, Pune, India Article Info ABSTRACT Article history: Received Feb 24, 2024 Revised Jul 30, 2024 Accepted Aug 12, 2024 This article presents the results of experiments on path planning and control of automated guided vehicles (AGV) using single, fixed ceiling mounted, monocular cameras and colored markers. The camera employed in the system serves as both a sensor and controller. Initially, the working environment is structured using colored markers for given applications. For every new setup, structuring the environment is essential. The image processing algorithm identifies the colored markers and their positions, which are then utilized for path planning and segmentation. The actuation time required to transverse each segment is calculated and then AGV is actuated accordingly. A transformation or inverse mapping matrix (M), predetermined, is employed for calculating world coordinates from given image coordinates. Path planning and AGV control are across various paths, both with and without static obstacles, in real-time applications. The colored marker detection and recognition accuracy for the given setup have been found cent percentage while the AGV reaches the goal point with an error margin of around 3.9% on straight paths, both with and without obstacles. Keywords: Automated guided vehicle Ceiling camera Control Mapping Path planning This is an open access article under the CC BY-SA license. Corresponding Author: Dr. Ramesh Pungle Department of Mechanical and Automation Engineering, P.E.S. College of Engineering Chhatrapati Sambhajinagr, India Email: rameshgpungle@gmail.com 1. INTRODUCTION Traditionally, in automated guided vehicle (AGV) navigation, it has been common practice to utilize a variety of sensors, both onboard (mounted directly on the AGV) and off-board (positioned on the ceiling or at suitable locations), for tasks such as path detection, obstacle detection, and navigation towards a destination while avoiding obstacles. In AGV navigation, the general tradition has been to use different sensors; on board (mounted on the AGV) and off-board (fixed at the ceiling or suitable location), to detect or generate a path, detect obstacles, and also to reach the destination by avoiding the obstacles. Onboard sensors are typically favored for their ability to detect and circumvent both static and dynamic obstacles along a given path. These sensors enable local path planning and motion planning by providing real-time feedback directly to the AGV. Off-board sensors, on the other hand, are often employed to plan the overall path, taking into account the positions of the AGV and static obstacles within the environment. However, employing multiple sensors onboard the AGV can introduce synchronization challenges among their outputs, potentially leading to complex control systems and increased response times. Alternatively, if a single off board sensor is capable of performing all of these tasks, the problem becomes significantly simpler to solve, and the complexity of the