Gear transmissions are widely used in vehicles, aircraft, ships, and wind turbines. It is of great importance to assess the dynamic mesh force or stress during both the design and maintenance phases. This paper begins by reviewing our previous work on identifying dynamic mesh force/stress using regularization techniques, vibration model-based approaches, and singular value decomposition (SVD) based Kalman filter (KF). To improve identification accuracy, a method based on Physics-Informed Neural Networks (PINN) is proposed to identify the dynamic mesh force/stress of gear transmissions. Additionally, we validate the performance of the proposed PINN model in identifying the dynamic mesh force/stress using testing signals.