<p>Machine learning (ML) has become the prevalent practice in the field of predictive modeling in mechanical systems, which allows the identification of performance patterns and detection of early signs of a malfunction. The use of MS together with conventional mechanical approaches like finite element evaluation has improved structural analysis and failure assessment. Furthermore, the ML has revealed fresh sets of material properties enabling the creation of new and unique materials since ML scans a vast database to uncover new potentialities. When we have appropriate specifications for mechanical material and apply ML algorithms to the design process, the correct combinations of elements and configurations are revealed. The paper also highlights the difficulties encountered when dealing with big and tremendously large data, the significance of preprocessing in training and how the various categories of training Learning; Supervised, Unsupervised and Reinforcement learning apply to material design. This work offers a great overview of the current use of ML in analyzing mechanical systems and material science and presents a possible trajectory for further development of the predictive modeling and material design.</p>