Magnetic position and orientation detection systems typically feature a permanent magnet that moves relative to a magnetic sensor so that the mechanical motion can be calculated from the sensor output. As a result of their reliability, precision and low fabrication cost, such systems are widely used in modern industrial applications like automotive gear shift detection, for wheel speed sensing and many others. Magnetic system design refers to the task of finding an ideal layout for such a system, to detect the desired motion in the best possible way. The design process can be a mathematically challenging high dimensional optimization problem and must account for multiple constraints and requirements that include stray field influences, ferromagnetic surroundings, limited installation space, desired resolutions, compensation of fabrication tolerances and many more. State of the art techniques for magnetic system design are discussed including analytical solutions, numerical methods like FEM as well as a novel design of computer experiments approach based on additive Gaussian process models.