This thesis focuses on the Mitsubishi MELFA RV-2AJ, a 5 Degree of Freedom (DOF) industrial anthropomorphic robotic manipulator and PUMA 560, a 6 Degree of Freedom (DOF) robotic arm, investigating the application of advanced computational techniques, specifically metaheuristic algorithms and deep reinforcement learning; to address the challenges of singularity and trajectory planning. In this paper, a detailed analysis and results of the algorithms have been presented along with comparative analysis of the conventional methods and DRL based algorithms. For calculating the forward and inverse kinematics of two robots, different approaches have been established but the motive was to cover different optimization algorithms to check which one is better in solving the problem for both robots. In case of trajectory planning and singularity avoidance, same DRL method has been established to verify the authenticity of the environment and optimized agent set by 6 DOF robot on any other robot of n degree of freedom.