Simulations are for optimization — of products and of processes. To be able to optimize designs, the foremost requirement is that simulations predict reality accurately. Secondly, speed of simulation is also important, because optimization usually involves running hundreds of simulations.
We use very diverse optimization techniques to solve your
design problems optimally. Relatively simple techniques are
search space methods such as gradient descent, response
surface estimation, swarm optimization and simulated annealing.
More esoteric methods, applicable in certain cases, involve
concepts such as topology optimization and Fréchet derivatives.