The asteroids (more precisely: objects of the main asteroid belt) and Kuiper Belt objects (more precisely: objects of the cold classical Kuiper Belt) are leftovers of the building material for our earth and all other planets in our solar system from more than 4.5 billion years ago. At the time of their formation those were typically 100 km large objects. They were called planetesimals, built up from icy and dusty grains. In our current paradigm of planet formation it was turbulent flows and metastable flow patterns, like zonal flows and vortices, that concentrated mm to cm sized icy dust grains in sufficient numbers that a streaming instability followed by a gravitational collapse of these particle clump was triggered. The entire picture is sometimes referred to as gravoturbulent formation of planetesimals. What was missing until recently, was a physically motivated prediction on the typical sizes at which planetesimals should form via this process. Our numerical simulations in the past had only shown a correlation between numerical resolution and planetesimal size and thus no answer was possible (Johansen et al.2011). But with the lastest series of simulations on JUQUEEN (Stephan & Doctor 2015), covering all the length scales down to the physical size of actual planetesimals, we were able to obtain values for the turbulent particle diffusion as a function of the particle load in the gas. Thus, we have all necessary data at hand to feed a 'back of the envelope' calculation that predicts the size of planetesimals as result of a competition between gravitational concentration and turbulent diffusion. Using the diffusion values obtained in the numerical simulations it predicts planetesimal sizes on the order of 100 km, which suprisingly coincides with the measured data from both asteroids (Bottke et al.2005) as well from Kuiper Belt objects (Nesvorny et al.2011).