Kriging, also known as a Gaussian process model, is a mathematical interpolation method used for simulation. The researchers in this study used Latin hypercube sampling to select input combinations for simulation, allowing for both uniform and non-uniform distributions. They discussed both deterministic and random simulation, which required adjustments in design and analysis. The study also covered sensitivity analysis of simulation models using functional analysis of variance, known as Sobol sensitivity indexes. Finally, the researchers discussed optimization of the simulated system, including robust optimization.