Solve your biggest product problems fast, sustainably and efficiently by Statistical Engineering.

Product problems are often treated as follows: problem is identified; some colleagues are working on it; an idea to solve this problem appears to be working; this idea is tested; in case of success the solution is implemented.

Very often ideas are tested that share the common error of not considering interactions within products. Hence tests are not successful. In case of failure next working cycle follows: failure is analysed, a new idea is generated and tested. Quite often dozens of working cycles have to be done. Problem solving is delayed work cycle after work cycle until an idea is successfully tested and appears to be satisfying. Design of Experiments (DoE) may uncover interactions.

However, experimental designs proposed by DoE specialists are often too complex. The number of experiments increases roughly by the power of number of variables. Colleagues do not understand too complex experimental designs in detail. They do not accept effort and practical failure risk of complex experimental designs.

Conmmonly, the successfully tested idea is not empirically secured, because widely used statistical methods are inapplicable for low failure rates at small sample size. Hence they are not used.

Thus in development, the subsequent validation is charged with hope. However, if the found solution fails during validation, a new engineering cycle follows with a new idea and test. After first failed validation a couple of engineering cycles is more rule than exception on the path of problem solving.

Later, if another problem occurs, it will be treated in the same way, because past problem solving did not add quantified know-how of functional relationships and root causes.

Even at quite low complexity this common treatment of problems is undependable, slow, and inefficient.

To solve problems dependably, fast, and efficiently, we recommend the early application of Statistical Engineering and Innovation Engineering. The number of working cycles until solution is cut by factors. The lead time to solution is significantly shortened. Capacity requirement is considerably reduced. Root causes and interactions are identified. The solution is proved and secured before validation with low number of samples and tests. Increased know-how of functional relationships eases solving of subsequent problems. With Statistical Engineering, problem solving of even the biggest product problems can be planned, measured, and controlled. Isn`t that amazing?