The damage identification process through Structural Health Monitoring (SHM) field has drawn extensive attention over the last decades for its numerous applications in failure prevention and maintenance decision-making. Several research in vibration-based methods for SHM have shown that a potential structural damage can be inferred from a change in the dynamic response of the structure. The aim of this paper is to detect and locate different damage scenarios in a benchmark bridge structure under a moving load based on Hilbert-Huang Transform (HHT). The data used in this study was obtained from the TU1402 benchmark towards enhancement of the value of SHM. The benchmark model consisted of a two-span steel bridge, where six levels of damage grouped in two damage region cases were introduced. In the proposed damage detection method, the transient vibration signals coming from a moving load in the bridge, are firstly decomposed into intrinsic mode functions (IMFs) using the Variational Mode Decomposition (VMD) approach. Then, the Hilbert Transform (HT) is applied to the IMFs. Lastly, the Marginal Hilbert Spectrum (MHS) and the Instantaneous Phase Difference (IPD) were used as damage indicators by comparing the undamaged condition of the bridge with each damage scenario. Results demonstrated that the proposed damage indicators were accurate for identifying and locating damage under transient vibration loads.