Sediment transport is one of the main morphodynamic processes in mountain streams. Understanding bedload transport is fundamental to prevent and mitigate hydrogeological hazards and for the resources management. The classic bedload discharge predictions are usually expressed by empirical formulae as a one-to-one relationship between sediment flux and water discharge. Unfortunately, this framework is often not sufficient to describe the variability of bedload discharge. In fact, large fluctuations in bedload transport rates are reported, even for steady flow conditions, over different time scales. Especially in case of mountain streams, the predicted and measured bedload transport rates may reflect differences of several orders of magnitude. Different processes, acting at different spatial and temporal scales, have been proposed as causes of the observed fluctuations, such as the intrinsic variability of particle paths, the migration of bedforms or the unsteadiness in sediment supply. During the second half of the twentieth century, some scientists started to study the bedload transport as a stochastic process involving the grains transported by a stream. The purpose of my research is to assess and improve the performance of a stochastic model in conditions representative of mountain streams and in presence of bedforms. To address this challenge, experiments will be performed in a narrow 2D flume, with steady flow conditions and stationary sediment feeding, for slope angles varying between 1 and 3°. The sediment feeding rate and the bedload tranport rate at the outlet will be measured using high-speed cameras and image processing techniques. Another high-speed camera will acquire side images of the flume for monitoring the evolution of the bed topography and for tracking the moving particles inside a control window. The analyses will be focused on activity and velocity distribution of the particles inside the control window. The collected data will allow the comparison of model outputs with the observed fluctuations of bedload discharge. The dependence of this fluctuations on different variables (shear stress, slope, antidune geometry, antidune migration celerity, etc.) will be investigated.