Gaining a better understanding of when electronics are about to fail can help engineers and companies be more effective in providing reliable electronics; this will also lead to a greater return on investments for businesses (Seggie, Cavusgil, & Phelan, 2007). Prior attempts of implementing algorithms have all failed due to the large bandwidth required to process the data in real time. The algorithms in this study will not require many resources and will use simple gate-logic and statistical data to alert an end-user of a potential failure. Scores, which are based on inputs from the end-user, for each of the four factors, will serve as the four independent variables for this study: (a) cost, (b) reliability, (c) component replacement, and (d) total system replacement.
Five different pieces of hardware will be subjected to the algorithms and tested five different times to validate if the prognostics are working. The prognostic-algorithms of years past are not intelligent enough to decipher large amounts of data. Van Bree, Veltman, Hendrix, and Van Den Bosch (2009) stated that rather than passively reacting to prognostic issues with limited fault-tolerances, the capabilities of a system should predict failures actively. The system should be able to reconfigure and control actions so that stability and acceptable performance of the entire system can be maintained (Van Bree et al., 2009). This research was intended to better understand the aspects of prognostics in the electronics world. This abridged literature review focuses on three key topics, which helped form and inspire the crux of the research questions: successful implementation of prognostics, time-to-market concerns, and technical risk.
Successful Implementation of Prognostics
Engineers have gotten better at diagnosing failures through embedded technologies such as boundary scan, BIT, and delta-scan. Nadeau-Dostie (2000) wrote that embedded technologies provide greater test access, higher test effectiveness, higher test coverage’s, and better fault isolation, which strengthens test applications. Until recent years, engineers still have not been successful in large-scale prognostics in the electronics world. Vichare and Pecht (2006) wrote that there has been an increased interest in the observation and monitoring of electronics; and systems in order to predict failures; and provide early detection to end-users to avoid major failures.
Carey (2011) led a project that examined outdated test positions, which were owned by the government and private electronic sectors. Several of the test positions contained instruments that are no longer found in current test positions. The author developed a medium to deploy embedded prognostics on legacy test positions. The test positions were studied to determine common problems across both sectors.
Masrur, Wu, Mi, Chen, & Murphey (2008) presented a case where prognostics were implemented in a power electronic switch for a direct current...