A physics-aware deep learning system is being developed to infer cellular traction from only the morphology/shape of the cell, which can be easily acquired by imaging. Cellular forces dictate cellular processes and the onset and progression of diseases such as cancer and asthma. However, the application of cell mechanics is hindered by the costly, time-consuming, and complex procedures to measure the forces. Currently, Traction Force Microscopy (TFM) algorithms face either an ill-posed inverse problem and discretization, which create problems including dangerously high sensitivity to measurement noise, or an even more complex experiment and extremely high computational cost. Besides, the current methods completely depend on in-vivo experiments, so cannot exceed the experimental parameter space. To resolves the limits of the current methods and to revolutionize the methodology in the field of cell mechanics, a generalizable physics-constrained deep learning system with adjustable physical and biological parameters is being developed to accurately and instantaneously infer and simulate the dynamics in cells and tissues under different conditions from merely the time series of cell morphology. Moreover, the application of this project exceeds cellular dynamics, as the fundament of the project is modelling physical problems with deep learning.