Statistics in Industry


Satisfied customers = product success.


Market-oriented concept – accurate planning – life-time equivalent design – qualified suppliers – development on schedule – sufficient validation – reliable product – satisfied customer.


That’s the theory.

In reality it’s often different.


Times are gone, when products and their components were developed elaborately and tested in high quantity to release them as technically mature as possible.


The tight global competition is a reason for shortening development periods and processes, high costs pressure leads to lean development and testing.


Nevertheless quality and reliability are more and more product key characteristics, so many industrial companies find themselves jammed between reduced development periods, costs and complex technical requirements.


This increases the risk to release products which are not mature for the market. Downtimes due to defect products make customers displeased and cases of serial damages may destroy a company’s profitability and its positive image – carefully built-up for years – in very short time.

But we think that’s not necessary!

We are convinced that integrating statistical methods over the entire product lifecycle can ensure highest reliability and quality, even with a rather small budget.

What does that mean in detail?

During the concept phase lifetime, reliability and availability targets have to be defined. In prototype phase, experimental plans help to find out the adequate amount of required components. This avoids on the one hand unnecessary testing and helps on the other hand to legitimate a certain minimum amount of samples.


Statistical analysis of measurement data is indispensable in order to be able to derive reliable statements from existing data. Thereby, e.g., the following questions will be answered:

  • How do our clients use our product in reality?
  • How does temperature influence the stability of components?
  • Do we have the production process under control?
  • What about our process capability?


Stochastic modeling is the basis for drawing conclusions from data.

  • Can we achieve our lifetime target with the existing material?
  • Which failure rate do we have to expect for the new control unit?
  • Which warranty costs will occur for the current product reliability?


Design of experiments is a central element in preparing and legitimating budgets and planning resources.

  • How many tests are necessary to assure the reliability of a component?
  • How long do tests have to be run at least?
  • How can we vary the adjustment of the experiment ideally?


If we have awaked your interest in this topic, we invite you to continue!


In Examples for applications we show you multifaceted fields of applying statistics.