There is a trend in agricultural engineering towards high-performance harvesting machines with growing operating width and throughput. As much as performance and throughput are rising, the transportation units are characterized by increasing transportation volume.

If harvesting and transport are combined in parallel operation (e.g. self-propelled forage harvester), the driver of the harvesting machine and the driver of the transport unit has to pay highest attention to the loading process. But losses of harvesting goods caused by missing the trailer have to be kept at a minimum, the complete transport volume should be utilised and collisions between the involved machines must be avoided. Loading processes with large-scaled machinery mostly imply that the visibility into the transportation unit is severely limited.

The main aim of the presented research project is to develop and analyse the potential of model based loading exemplified on a forage harvester and a corresponding transport unit. The forage harvester had been used as the prototype for a GPS-based position control of the spout. The model based loading means an enhancement of the automation of the loading process. First objective of this research project and the main focus of this presentation is the development of a software model of the heap of bulk goods. Basal analysis on heaps of agricultural goods like grass and maize silage are essential. By combining the software model, the space model of the transportation unit and the throughput, the current status of loading is predictable and different loading algorithms can be spotted, tested and scrutinized with regards to efficiency and the facilitation of work.

The research project started in 2007 and will run for 3 years. Besides the general setup of the project, the software principle concerning the bulk heaps and the loading state will be presented as well as the latest project results.