The future of Reliability Assurance

I recently attended an excellent Reliability Seminar and Workshop at the University of Greenwich (14-15 September) http://reliability-outreach.gre.ac.uk/. A wide range of topics was covered including Software Assurance, MEMS, and Six Sigma Techniques. For the purpose of this article I will concentrate on the subject matter relating to Reliability Assurance.

I have previously emphasised the importance of Design for Reliability, and how it is important to have an overview of Reliability issues throughout the product development cycle. I have also made reference in previous writings to the “mission profiles” for application conditions as defined in standards such as JESD94A.
An underlying theme to the Greenwich workshop was consideration of environmental conditions a product experiences in the real world.

First of all consider the journey that a consumer product such as a laptop makes on the way to the customer. We might imagine that dispatch by air is a safe option, but think for a minute. How often have you watched baggage handlers from the window of an aircraft as they throw packages across the tarmac?
Next consider the operating environment of the product. How close is it really to the “mission profile”? Real measurements of “power on hours” and ambient temperature conditions may vary significantly from the assumed profile. We may also be underestimating the amount of mechanical stress the product receives in everyday use. The only way to be sure of real world conditions is to make the relevant measurements.

Moving on to the Reliability Qualification, we also need to question the relevance of the standards to which the product is being qualified. Historically the focus of Reliability activities has been on the Reliability Qualification itself, probably because a successful qualification is usually a legal requirement. However we only have to consider the numerous military handbooks cancelled during the 1990s and the draconian nature of the Telcordia standard (nothing will pass all of it !) to realise that to blindly follow a standard is ridiculous. It is far better to agree an appropriate programme of work with the customer in advance.

At the end of the day we all want to guarantee the Reliability of a product in the field. Regardless of whether we have “passed the qualification” the ultimate goal is to meet the product warranty and to achieve as low a level of field returns as possible. In my view we should be thinking in terms of Reliability Assurance rather than Reliability Qualification.

There are many components to Reliability Assurance. In my view this includes the various DfX activities (DfR, DfT, DfA, DfM, DfY etc.) which begin at the study phase, review of existing Reliability Data, and a relevant Reliability Qualification based on a verified real world environment.
The Greenwich workshop completed my personal model for Reliability Assurance, and underlined my belief that we must consider the entire product creation cycle (and its life in the real world) to have a complete understanding of product reliability.

References : Reliability OutReach Workshop, University of Greenwich, 14-15 September 2009.
Specifically the following presentations:
Reliability Prediction Methods : What We Should Use and Not Use - Dr.Michael Pecht, University of Maryland
Compliance is not enough - Bob Page, Reliability Plus
Product Reliability – Managing the Time-Bomb - Dr Nihal Sinnadurai, ATTAC