Thesis & Research

Authors: A. K. J. Saudagar, S. S. Swapnil, S. K. Sarker,     A. B. Dibya, M. T. Islam, M. A.–T. Roni, H. A. Al sagri, and K. Muhammad

Invention Details: In this invention, our system is designed for precise speed control and reliable track adherence, using real-time line type detection to anticipate and adjust speed, even at high velocities. A novel data generation module produces diverse training images for a Support Vector Classifier, enhancing adaptability across track conditions. The system features an advanced adaptive image processing technique, optimized for small-scale microcontrollers, ensuring accurate line type prediction under different lighting conditions. A unique speed calibration method calculates base speed using the surface’s frictional constant and IR sensor data. PID parameters adjust dynamically based on the generated error from the current position of the IR sensor array for smooth navigation on both straight and curved tracks. Integrating Herringbone gear and advanced bearings provides robustness, along with silicon tires which ensure superior traction and durability, boosting system robustness for high-performance applications.