This project has been implemented under the supervision of Dr. Ashish Karn
The importance of Autonomous mobile robots is increasing for individuals and industries, enriching the automation and robotics sector. Various algorithmic strategies have been incorporated into mobile robotic technology, allowing experts to use it while also making it famous among novices. However, implementing such algorithms on a robot to perform automated tasks can be expensive and complex, as multiple sensors will be required for gathering data, and the high price of multisensory systems. This project provides an algorithmic, and cost-effective solution by developing the optimized algorithm, using multi-ultrasonic sensors compared to other expensive sensors, implemented within the terrestrial robot. The proposed approach is based on four fundamental processes: sensing, localization, map generation, and humanoid visualization. Three ultrasonic sensors (HC-SR04) are positioned in three distinct axes on the terrestrial robot, collecting data points from obstacles in three different directions within the linear based arena. The acquired data points aid the terrestrial robot in creating its local environment map in this new method. The terrestrial robot navigates confidently via the achieved linear path using a variety of range data. The data set is then utilized to visualize the humanoid mapped environment. Experimental results have shown that our optimized algorithm is quite efficient with way less-expensive multi-ultrasonic sensors, making it a very cost-effective hardware solution for real-world automation problems. .
Keywords: Terrestrial Robot, Robotics, Computer Vision, Deep Learning