Technology Product Selection Application: 

Various vendors offer technically suitable proposals. Adding reliability (MTBF) analysis to the evaluation criteria allows managers to anticipate differences in quality and include them in the analysis. 

There are three major improvements to analysis that are made using reliability.The first is to make sure that the new products are at least as reliable as the existing. This is not always the case, so a baseline of current products is essential. Second, the products should be ranked on the basis of reliability, as if there is a significant difference in reliability managers should always select the best. Price and reliability do not currently correlate. Third, the long-term total cost of ownership should be calculate to determine if a premium price product (including service contract) is worth the additional investment.

Example: Total hardware cost ranges between $ 2.5 million and $3.0 million for the big name brand. ( Competitive bids often result in a spread of 20-25% between offerings.)   If the lower cost product is at least as reliable as the higher priced product, savings is $500,000.  But if all products were roughly the same in terms of MTBF, then the analysis can shift to including the total cost of ownership including the total cost of the service (maintenance or warranty uplift) for the intended useful life. Most service plans are high profit items and the range of discounting rarely reflects the realistic costs to perform service. A savings of 30% (or more) should be anticipated using reliability data in a service contracting application.

Smart Meter Test Application: 

Regulators have approved a pilot test of a combination of equipment and rate design for a limited period. Measurement of consumer reaction is a high priority, but the equipment itself is also under test. Beginning with the first device, the volume of installed devices can be used to calculate the MTBF of each component so that engineering specifications can be validated and early stage (infant mortality) issues resolved quickly. All new products can be compared to each other, as well as to the baseline (original) environment to calculate the total cost of ownership for the fully deployed environment. These projections inform the business case analysis required by regulators to approve subsequent deployments.

Example: Experience tells us that using current technology, each new meter in the field will cost twice as much to repair as to install. So for a unit cost of $250, the repair costs of a truck roll to repair or replace the unit will be roughly $ 500. With meters expected to last ten years, current failure rate data suggests that during that period anywhere from 50% to 100% of all the initially deployed meters will have a service event. Every failed meter costs $ 500 - so a very small reduction in failure rate has immediate payback. Using test and pilot data to discern small differences in quality between products is essential to avoiding potentially catastrophic costs in a production deployment.

Parts and Service

Many products remain in the field long after the vendor ceases manufacturing the model. Looking exclusively to the vendor warranty to manage services issues is invariably inadequate over the long term. Reliability monitoring is ideal for managing long-term service issues. MTBF of the device as initially deployed immediately shows the types and frequency of failures. Parts and spares can be stocked appropriately, and service teams hired and trained to manage the specific problems. Over time, changes in the failure rate will guide both vendor and user to develop a long-term plan for service after the warranty period. Some of these options may include depot repair, advanced purchases of spares, used equipment purchases for parts, and even independent service contracts.