Physics-informed neural networks (PINNs) [1], [2] incorporate physics-based knowledge (such as displacement and force boundary conditions, governing equations) in neural networks (NNs) through loss functions, presenting potential for the development of robust digital twins of physical systems and processes. Despite progress achieved with PINNs [3], a key challenge currently is in developing PINNs that are capable of representing the physical system under variable input, i.e., spatially and temporally variable loading in the case of civil and mechanical structures. This paper presents a novel PINNs model that is trained and validated to predict the response of bridge girders under a moving load. In particular, the efficacy of the PINNs model is assessed for virtual sensing with real-world monitored data – virtual sensing refers to scenarios where the model is applied to responses (e.g., strain time-history diagrams, deflection, and internal forces) at locations that are not physically equipped with sensors. The PINNs demonstrate the capacity to capture the girder strain time profile in scenarios with and without monitored data, demonstrating consistent trends and good agreement. [1] M. Raissi, P. Perdikaris, and G. E. Karniadakis, “Physics-Informed Neural Networks: A Deep Learning Framework for Solving Forward and Inverse Problems Involving Nonlinear Partial Differential Equations,” J Comput Phys, vol. 378, pp. 686–707, 2019, doi: 10.1016/j.jcp.2018.10.045. [2] M. Raissi, P. Perdikaris, and G. E. Karniadakis, “Inferring Solutions of Differential Equations Using Noisy Multi-Fidelity Data,” J Comput Phys, vol. 335, pp. 736–746, 2017, doi: 10.1016/j.jcp.2017.01.060. [3] A. I. F. Al-Adly and P. Kripakaran, “Physics-informed neural networks for structural health monitoring: a case study for Kirchhoff–Love plates,” Data-Centric Engineering, vol. 5, p. e6, 2024, doi: DOI: 10.1017/dce.2024.4.