Industrial settings require autonomous navigation to automate dangerous, time-consuming, or difficult tasks. Avoiding obstacles is crucial to autonomous robot safety and effectiveness in complex and changing industrial settings. Because of uneven ground, poor grip, and limited sensor data, rough terrain is difficult to navigate. This paper shows how deep reinforcement learning helps industrial robots avoid obstacles. The goal is to create a solid framework that lets a robotic agent autonomously travel from a starting point to an endpoint, avoiding obstacles. The suggested method uses reinforcement learning algorithms and deep neural networks without affecting each other. The computer agent learns how to plan its path and avoid obstacles. The study involves building simulation environments with MATLAB Simulink and planning experiments to test the framework. The manuscript also discusses hardware limitations, real-time operations, and deep reinforcement learning model challenges. This manuscript advances industrial robotics by providing a practical obstacle avoidance solution using twin-delayed deep deterministic policy gradient (TD3) and Deep Q-Network (DQN) deep reinforcement learning algorithms, which could revolutionize manufacturing and other industrial tasks.