Evaluating Warehouse Automation in a High-Mix Manufacturing Context at ZF Padova S.r.l

When a factory assembles products in many configurations and in few units of each, the warehouse that supplies its lines must hold a large number of codes, each requested rarely, and deliver them when the line calls for them. Compact automated warehouses, adopted above all in distribution and e-commerce, promise to do this in less space and with less handling. Assessing them in a factory, however, requires knowing what capacity and service they must provide, information that a manual warehouse seldom collects and that evaluation models take as already available. This thesis addresses the problem at ZF Padova S.r.l., where four manual warehouses feed the marine transmission assembly lines and the company plans to replace them with an AutoStore system, the solution presented by Element Logic. Using one year of operating data, the work reconstructs the load handled today and derives the performance the automated system must deliver, then sizes the system with published models and without drawing on the supplier's calculations. The result undergoes a sensitivity analysis and is compared with existing installations and with what the integrator proposed in a first meeting. Since the cost is not yet known, the economic assessment identifies the maximum investment that the savings could pay back. The system must above all hold many codes in a small volume rather than sustain many movements, so the installation is small: two ports, a few robots and a grid covering under a quarter of the floor area used today, dimensions that the integrator also reaches. The calculation rests on mostly prudent assumptions, although the service to the lines in the busiest hour still has to be verified. The economic case remains open: the only saving timed so far, moving material from the upper level to the picking level, covers a fraction of the expenditure indicated by the supplier. What matters for the decision is how much handling work the installation would eliminate, a figure that ZF can obtain right away by timing the picking operations carried out today. Once the offer is received, the data and the sizing developed here provide the basis for checking its individual items and estimating its return with the actual price.