In rock mechanics and rock engineering, numerical analysis techniques have become an essential method. However, numerical analysis requires high computing costs and has the disadvantage of difficulty in directly integrating measurement data to make model predictions through data assimilation. Physics-Informed Neural Networks (PINNs), which incorporate partial differential equations (PDEs) as components of neural networks, have been actively used to address these issues. In this paper, the verification of PINNs for simple PDEs based on geomechanics was conducted to assess the applicability of PINNs. PINNs were compared with traditional numerical analysis methods, and the performance of PINNs was verified for linear elasticity and consolidation equations. This verification process included two approaches: deriving the solution to the PDEs as forward analysis and estimating PDEs coefficients through inverse analysis. This research explores the potential of PINNs in modeling complex multiphysics phenomena and enhancing their practical applicability.