This thesis addresses the storage-retrieval problem for a fully automated storage facility. We alter an existing deterministic and a factor-based storage and retrieval model to be applicable to a fully automated multi-deep storage facility. The models are able to handle uncertain supply and variable demand. Some practical attributes are added to the class based models, which can be inuenced by the assignment of penalties. The results show that the model based on deterministic supply shows better results than the factor-based allocation model, when applied on a multi-deep facility.