Statistical distribution – it has important role in general life and of course in engineering specifically. So, what is all about?
During morning commute either by bus or vehicle, you take some budget of time. From your daily routine experience, you with very high probability, can predict arrival time. You also know how much time it will take, in less ordinary cases, like slippery roads, or cars accident along you route.
If some will ask you question, how much time take you reach your destination, typical answer will be following:
“Usually it takes me 25 minutes, when there is no traffic it can take top 15 minutes, and if there is traffic, it can take 55 minutes.”
And this description is exactly how distribution work. In average (most of the chances) – 25 min, in special cases it can vary between the 15 to 55 minutes.
If we will build a statistical distribution (assumption that is follow normal distribution) it will look like that:
Returning to our previous example of dishwashers the target time was 11,500 hours. After running a test to failure, can be generated a distribution based on time to failure and it will have 3 possible outcomes:
Case A – Majority of dishwashers will fail long after target time, probably customers will love this option most, dishwasher will serve longer period, than declared. For selling company of those dishwashers it is an opportunity of cost reduction.
Case B – Some of dishwashers will make, it some will fail before the target time, and here it is important to get understanding how many samples, company will get back due to customer complain.
Case C – A night mare for company and for customers. Majority of dishwashers will fail before target life.
Generate a distribution can be helpful in prediction of life cycle of the product and can show honest picture to higher management.
Leave a Reply