This work aims to perform optimization of manufacturing for the three products, the shaft, the axle and the gear, and training of neural network for the three buffers. Purpose of the optimization process is to find the optimal number of machines, reduce the cycle time, increase the level of machine efficiency and to find optimal number of buffers with the Experiment Manager tool. Experiment Manager tool shall be applied with neural network in the Tecnomatix Plant Simulation and calculate the simulated behavior of manufacturing process. Introductory chapter briefly specifies the paper s subject referring to the optimization of manufacturing process wherein three different products are obtained. Artifical intelligence in the expert system field and neural networks has been used in the optimization process. It has been shown that neural networks are a specific group of artificial intelligence, and the paper also considers some of the neural networks mostly applied in practice.