SRC Forum - Message Replies
Forum: Reliability & Maintainability Questions and Answers
Topic: Reliability & Maintainability Questions and Answers
Topic Posted by: Reliability & Maintainability Forum
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998
Posted by: Paul Higgins
Date posted: Sun Sep 11 15:47:35 US/Eastern 2005
Subject: Field Data
I am analysing field data from a piece of electro pneumatic equipment. The dat however refers to aircraft flying hours and not necessarily equipment operating hours. When a piece of equipment is returned for repair we record equipment operating data, however there are in excess of 1000 systems in service and we have only had 240 failure in 9 years.
The problem is because the operational time of the population is not known it is not easy to determine the true reliability of the system and the equipments reliability is a contractual requirement.
Does anyone have any thoughts on the best approach to take with the customer or any tips on analysing this type of data? We are trying to work with the customer to determine the average equipment operating rate during a known period of flying but as yet this is not known.
Subject: Convert calendar reliability to operating hour reliability
Reply Posted by: Larry George
Organization: Problem Solving Tools
Date Posted: Fri Sep 23 19:20:06 US/Eastern 2005
Please refer to http://www.fieldreliability.com/Nwslt5.doc page 4 for "Calendar Age<->Operating Hours." It deals with both sell-through time and with random operating hours per unit calendar time. It says, "People who should know better rescale the calendar age variable by the average operating hours per calendar age unit." [Sorry Bruce, MTBF*(op hour/calendar hour) is correct only for constant (op hour/calendar hour).]
Use the ->distribution<- of random operating hours per calendar hour, not its average, to convert from calendar hour reliability to operating hour reliability. [The same advice applies to a subsystem whose operation is not the same as the system operation.]
Gather a random sample of operating hours per calendar hour. Estimate subsystem reliability per calendar hour. Transform subsystem reliability per calendar hour to subsystem reliability per operating hour using the distribution of op hours per calendar hour. If you want an example, send data. I will send back nonparametric reliability estimates with calendar time scale, distribution of op hours per calendar hour, and nonparametric reliability estimates with op hour time scale, free of charge. email@example.com