Q 80. A component manufacturer (OEM supplier for a complex finished product) faces several unique challenges. One of them is as follows -  

 

The field failures reported by the client result in repairs and replacement costs to be borne by the component maker. Let us assume that the failures reported by end customer and client are genuine. Client wants improvements to be made so as to reduce reported component failures.

 

Let us assume there are 20 possible reasons for failures. The component manufacturer wants to make improvements but has no mechanism to validate the root cause of field failures. This is due to two reasons. 

 

1. Failures show themselves only after usage in field over many months.
2. Accelerated life testing is very costly and doing hypothesis tests for each cause is cost prohibitive. 

 

Improvements (based on guessed root causes) made today will show their success or failure in months or years to come. How should this situation be handled by the component manufacturer? 

 

Note for website visitors - Two questions are asked every week on this platform. One on Tuesday and the other on Friday.

• All questions so far can be seen here - https://www.benchmarksixsigma.com/forum/lean-six-sigma-business-excellence-questions/
• Please visit the forum home page at https://www.benchmarksixsigma.com/forum/ to respond to the latest question open till the next Tuesday/ Friday evening as per Indian Standard Time.  
• The best answer is always shown at the top among responses and the author finds honorable mention in our Business Excellence dictionary at https://www.benchmarksixsigma.com/forum/business-excellence-dictionary-glossary/ along with the related term.

 

 

 

In my opinion the OEM should follow the following steps:

 

1. List the 20 possible reasons for failures as three lists arranged in:

• decreasing order of criticality (i.e. the most probable root cause)
• decreasing order of probability of occurrence
• increasing order of cost to mitigate

(The  possible reasons could be identified based on a detailed Fault Tree analysis considering every possible cause, even the remotely possible ones , Cause Effect diagrams, fish bone Analysis and so on by a task force team comprising of experts)

 

2. Focus on the most probable cause or causes with high probability of occurrence and either mitigate or eliminate them, keeping in mind the cost involved and the investment that the business can afford.

 

3. Use past complain reports and analyze average time period after which the components have failed. Based on this  analysis they should come up with a conservative estimate of the average life of the component.

 

4. Proactively replace the components in field, that are nearing the average life period, with new components.  ( i.e. before the actual failure causing downtime)

 

5. While installing the new components set up mechanisms ( instruments, software checks) to  trap and capture data when failure occurs next time to verify real reason E.g. temp sensors recording data to see if a temperature rise is causing issues or a vibration sensor to check for intermittent excess vibrations and so on.

 

6. Test and analyze the current status of components removed from field for any failures.

 

7. When steps 4,5 and 6 are repeated for 1 or 2 consecutive cycles there is a high chance that the real cause of failure can be identified and rectified.

The given situation is that of a reliability failure, where time is a factor. Obviously it is the infant mortality rate that is causing pain to the client.

If the option of accelerated testing is ruled out due to cost considerations, the following approach may be adopted for quick identification of the highly probable causes.

1. Assuming that sufficient failure data is available, plot the failure rate vs time graph. This will usually tend to take the shape of an exponential distribution, with high concentration of failures in the early periods.
2. From this plot, determine a time period beyond which the failure rate tapers down to a safe levels.
3. Pick up reasonable number of samples of failed components from the “early failure period” and seek for equal number of samples of components that are successfully performing even after the ‘safe cut-off’ period. (This will need a client co-operation as well as willingness by the supplier to replace those good components, to support this exercise)
4. Now we have a set of “survived components” and a set of “failed components”.
5. Depending upon the type of component, list out a set of quality characteristics that may be compared between the “survived components” and “failed components”.
6. The sets of observations for each characteristic from the “survived” components need to be compared against the corresponding set from the “failed” components.
7. To decide on significance of the differences for each characteristics, appropriate hypothesis tests may be applied where relevant.
8. As a result of this exercise, the supplier would be able to re-define certain specification tolerances and manufacture components that are certainly bound to be more reliable.

The other alternative / supplementing approach could be to collaborate with the client for sharing investment on accelerated testing. If setting up such facilities are not feasible, the services of external laboratories may be sought. Ultimately, the success is going to be win-win for both parties!

In case of this situation where the improvements are made at a later stage of manufacturing, the best way is to wait for the results based on the life of components and work on a sample to find the failures. Based on this result, it would be great to provide a replacement of part/component voluntarily instead of waiting for each of the customer to complain in future. This can be done to components only with high failure or high risk of failure. Meanwhile alternate testing methods like reliability testing can to be used to find out the results of improvements applied so that the manufacturer has enough inventory held while the actual failures occur in future. This would atleast minimize the damage of not having enough inventory available to meet the customer demand

Venugopal has mentioned a study of failed components versus survived components with the same/ similar used life. This becomes an extremely important thing to try. This also seems to be the only low-cost possibility in the given circumstances. His response is selected as the best answer.