Most likely the greatest barrier to improving semiconductor or product reliability may be the lack of understanding of what it actually means. A lot of the resistance you can expect to implementing reliability can come from people who believe that quality and reliability are exactly the same thing.
Quality is conformance to specification, while reliability testing is quality in the long run. Reliability testing from an individual perspective is that “the product works the way it absolutely was likely to work for its desired period of good use.” Before starting improving reliability, you should be alert to what you are actually improving.
You’re going to be amazed at what goes on within the organization once internal barriers commence to break up. This phenomenon is similar to compared to a long‐term investment in a retirement account. At first, the retirement investment is small plus the amount routinely contributed does not may actually enable you to get any closer to that final goal. However, in the long run this amount becomes significant since it begins to grow exponentially. The exact same effect is going to be seen in the company after the organization begins to start to see the benefits of testing for product reliability.
There are two necessary requirements which can be needed to deliver the knowledge of the method. The first requirement will be have a charismatic leader or champion, who is highly experienced in the method and that can communicate clearly what needs to be done. Salesmanship here could be of good value. The 2nd requirement is that management has to offer the necessary changes necessary to implement the new process. Both these requirements are essential ingredients for success. Management support particularly is a total necessity.
The 2nd necessary ingredient in the implementation process is commitment. The commitment needs to be in manpower, capital resources, schedule allotment, plus in management. The management commitment must result from the highest level. In addition, the commitment should always be an element of the five‐year planning, since the first few years the payback may not seem apparent. Once senior management has made a decision to implement product reliability testing in to the organization, a gathering should really be planned with middle management and outside consultants in which shared goals for implementation could be set plus the foundation for implementation established. Some companies use weekend retreats with this meeting. The utilization of reliability testing on electronics really should not be seen as an experiment.
The businesses which do have product reliability testing programs think about this included in their core competency. From FMEA, HALT (highly accelerated life test), HAST (highly accelerated stress test), and from previous programs, the lessons learned can be used to future products. The lessons learned can be captured in a database and made available to everyone through a computer‐based retrieval system. Since these databases can get quite large, there should be the search engines effective at finding studies centered on key words or subjects. Second, lessons learned should always be summarized into a design for reliability guideline this is certainly updated and communicated to the design community.
In most case, the brand new reliability process is on a critical path and photonics reliability equipment needs to be considered. Resistance to you can easily mount, because early market entry is really important for profits. Any added activity delays market entry. Sometimes, as an alternative, management chooses to place the reliability process on a sidetrack so that the existing product development is not slowed up by anything, especially the untried reliability processes. Placing the reliability process in parallel towards the regular product development flow could work, however the results will be very small in addition to savings in development time won’t be realized. Later, the evaluation of this new reliability process could have earned little support, because the data to guide continuing the reliability effort will likely be almost nonexistent.