Browse Prior Art Database

System and Method for using component Refurbishment status in Isolation, Repair, and Failure Analysis

IP.com Disclosure Number: IPCOM000127417D
Original Publication Date: 2005-Aug-29
Included in the Prior Art Database: 2005-Aug-29
Document File: 3 page(s) / 35K

Publishing Venue

IBM

Abstract

The refurbishment of components that are returned from field system maintenance is common practice in the electronics industry. No refurbishment process is one hundred percent accurate and the potential for escapes exists. The potential from intermittent or latent faults that are detected in the refurbishment process makes the probability for a failure of a refurbished component slightly higher than a component that has not exhibited a failure previously. Disclosed is a method for accounting for this probability in real time recovery and isolation, system repair, and failure analysis for returned components.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 32% of the total text.

Page 1 of 3

System and Method for using component Refurbishment status in Isolation, Repair, and Failure Analysis

When certified service parts are used in systems, it would be of value for the system error recovery and component isolation programs to use that information in analyzing component failures in determining the components that are called out for repair. Though certified service parts have been through a verification process to certify them, latent or intermittent failures can allow components to pass through a screening process. The information that a certain component has previously been identified as failed and has been replaced in a service repair can be used to weight error recovery and component isolation analysis algorithms to help identify the parts that require service.

This disclosure uses persistent data storage on a component that is accessible by the system error recovery and component isolation programs to identify a component as a certified service part. This information is used by the system error recovery and component isolation programs to weight problem analysis in identifying failed components. This can be especially useful in repair scenarios where there is a 50/50 probability for two components during a failure scenario.

All new components will be provide a persistent storage location that provides an indicator of the certified service part status of the component. The persistent storage shall be accessible for the system operating code to retrieve the status of the certified service part status of the component. The data stored within the persistent storage location will provide information that describes the component status with respect to certified service part (true = yes, component is a certified service part : false = no, NOT certified service part). As an optional extension, the certified service part status information could also contain information regarding the number of times a given component has been processed as a CSP. This field could be generically referred to as a CSP counter.

During system operation the system error recovery and component isolation program will collect the certified service part status information for all components. During system error scenarios where analysis is conducted to determine the components to be isolated or fenced by the error recovery and component isolation program, a weighting factor will be established that accounts for the involved components certified service part status. Depending upon the component and failure modes the weighting factors applied will have to be developed uniquely for each component.

As a generic example; if component 1 has a certified service part status of true and component 2 has a certified service part status of false, then when calculating the certified status part weighting factor the following logic is used. For component 1, with CSP status = true, the CSP weighting factor would be a value that increases the probability that the comp...