ROTATING IR SENSORS FOR ROBOTIC MAPPING AND NAVIGATION

Low cost infra-red (IR) sensor is widely used in many robotic applications such as obstacle avoidance, path finder and shape detection. LIDAR is high accurate and precise to be used in many applications. However, it is too expensive for a simple robot project. Therefore, IR sensor is proposed to be applied for a mapping and navigation, however with reduced accuracy. The main focus of conducting this project is to develop a rotating IR sensors that can be operated accurately in object detection. The prototype uses a rotating IR sensors which able to scan 180° of the surrounding area with 15° step angle. The prototype has been installed on mobile robot. And then, the mobile robot has been programmed to move 3 m x 3 m in known environment filled by random obstacles. Therefore, it is necessary to develop a prototype that has acceptable sensor reading to create a 2D map. Two different IR sensors have been tested on distance measurement and result are the accuracy varies at different object points. One thousand distance measurement have been recorded against known object at each point with 10 cm increments. The accuracy for both sensors become lesser, as the distance of the object increase. Nevertheless, the reading of the sensors also affected by the ambient light of the sensors as well. Relative frequency and average distance for each 10 cm increments have been analyzed to obtain more accurate results. Average distance from 20 readings for each step angle is used to represent the sensor reading. Sensor readings have been recorded in Excel wirelessly and 2D map has been created by Matlab. The results obtained from the testing on the mapping of the surrounding area have proved that the system works fine on the 2D mapping, however necessary improvements should be taken on the accuracy as well as the sensitivity against the objects detected.