The Inventory Movement Optimization Problem (IMOP) is a new scheduling problem arising in industry. A set of N pallets has to be prepared with different products in a processing area with T preparation slots (T < N). At every step of the process, pallets of products of a single type are requested from the inventory, and brought one by one by an AGV. When a pallet of product arrives in the processing area, a palletizing robot feeds the pallets where the product is requested. The pallet of product is then returned to the inventory, and as soon as a pallet is completed, it is transferred to the shipping area. The next pallet is inserted in the available slot. The objective is to minimize the number of movement of the AGV from the inventory. We develop a MILP model and solve it with a commercial solver. Comparison is conducted with a competitive heuristic and with an implementation on LocalSolver on real instances recovered from a client of company Fives Syleps.