Collective phenomena, when a large number of individual units often with very simple pairwise interaction display complicated behavior, are omnipresent in our life. These systems range from physical systems (molecules, granular materials) to the natural world (evolution, herding) and down even to human society. One of such systems is the multitude of pebbles found in rivers, lakes and shores. Pebbles start their life as fragments in the mountains, or shores, but then the work of the river, flood and storm transform them to smooth nice pebbles. This erosion process is the result of billions of collisions which form the size and the shape of the pebbles. Depending on the beach one may find only pebbles with similar size and shape while in other places the size distribution of the pebbles is much wider. We believe it is due to the nature of the mechanical process which eroded the pebbles. The goal of this PhD work would be to identify the relevant mechanical processes, use discrete element simulation to study the collision of the particles of different size and shape and finally determine the evolution of the particle size distribution. The student may also participate in experimental work. If this work is successful then a binary collision kernel can be established for this stochastic process which can be solved using the machinery of nonequilibrium statistical physics. The work would be done in the scope of the MTA-BME Morphodynamics research group.