Real-time displacement monitoring of earth crust and engineering structures with the Global Navigation Satellite System (GNSS) is highly important for detecting and preventing the damage and possible loss of life during natural phenomena. Nowadays, the Real-Time Kinematic Precise Point Positioning (RT-PPP) method using a stand-alone GNSS receiver can provide fast and accurate dynamic displacement. Trimble Real-Time Extended (RTX) technology enables real-time precise positioning with cm-level accuracy in a short convergence time. The present study investigates, for the first time, the dynamic displacement detection performance of RT-PPP based on CenterPoint RTX correction service for early warning systems and rapid risk assessment. To evaluate the performance of Trimble RTX technology, a series of shake table experiments were conducted and harmonic oscillations with different frequencies and amplitudes were generated. Trimble RTX solutions were obtained at 20 Hz sampling rate and compared with Linear Variable Differential Transformer (LVDT) data, considered as reference, in both frequency and time domains. Results show that Trimble RTX technology can accurately detect the frequency of harmonic motions, whose amplitudes are slightly different from LVDT values because of GNSS inherent noise. The average of Root Mean Square Error (RMSE) values obtained from the differences between LVDT and filtered RTX-derived time series for all events is 3.6 mm. These results indicated that the CenterPoint RTX technology could be employed for real-time and near-real-time applications to monitor structural and earthquake-induced motions accurately.