Page 100 - tech fest 2025 אשדוד חוברת תקצירים
P. 100
100
Path Planning and Obstacle Avoidance Using Reinforcement Learning in a Differential Drive Robot
ME-C-33
Yuval Levy; 859yuvyuv@gmail.com Jonathan Peretz; ISR400@gmail.com
Advisors: Dr. Itzhak Levy1, Mr. Yogev Attias1 1SCE - Shamoon College of Engineering, Ashdod
This project developed a reinforcement learning-based control system for path tracking and obstacle avoidance in a differential drive robot. Our robot is equipped with independent front and rear steering mechanisms, enabling improved maneuverability and control. Using a Python-based simulation environment, this system autonomously refines its driving policy by learning from environmental interaction. The final implementation demonstrated accurate and efficient navigation, while highlighting the practical potential of learning-based control in autonomous robotic systems.
Keywords: autonomous driving, differential drive robot, path planning, reinforcement learning, robot navigation
