Random-Tandem Queues and Reliability Estimation, WIthout Life Data

WIthout LIfe Data??? Do you estimate and track the field reliability of your products and their service parts? Life data include ages at failures and survivors’ ages, or births and deaths, by serial number or name. Is that reliability information worth the cost of life data? Is there an alternative to life data besides AFR (Annualized Failure Rate = failures/installed base)? If you do not know the age-specific field reliability of your products or their parts, consider this alternative. Estimate nonparametric field reliability and failure rate functions, without life data, from financial accounting data.  Life data is statistically sufficient but not necessary to make nonparametric population estimates of field reliability, survivor, and failure rate functions! Generally accepted accounting principles require financial accounting data, without invading privacy and without unwarranted distribution assumptions.

Financial accounting data includes revenue and cost data that contains “ships” and “returns” counts. Ships are births, production, sales, or installed base. Returns are deaths, complaints, failures, repairs, and even spare parts sales or usage. Ships and returns counts are available for service parts too. Bills of materials contain parts’ counts. They allow conversion of product installed base to parts’ installed base [gozinto theory].

Ships and returns counts are statistically sufficient to make nonparametric estimates of field reliability, survivor, and failure rate functions, without life data and without unwarranted assumptions. They are population data and are free. Sorry, it takes some work to obtain ships and returns counts from revenue and service costs. However, it is easier, cheaper, and more accurate than collecting life data, especially a sample of life data.

What is the value of nonparametric field reliability and failure rate statistics informationwith vs. without life data? The link below is the book comparing alternatives and describing what to do, without life data. It is first draft, bound to change. I will upload revisions now and then. Please let me know of criticism, errors, and questions, pstlarry077@gmail.com. Thanks.