Experimental data were used to develop a statistical model of mixing for the in-feed system to the grader, and a separate model for the mixing on the packing lanes. Appropriate model design was contingent on the results of the mixing, but the expected issues with developing the model had some implications on the experimental design.