I make the initial observation that all of Mr. Li's suggestions must be viewed in light of the fact that we do not currently know what type of facility that DJC may construct, or indeed if it will construct any facility at all. This uncertainty may have consequences. Unlike DJC, we purchase much of our machinery from outside suppliers who have in the past not only disclosed our intentions to competitors, but actually sold them the same products so that they could more effectively compete with us. In addition, DJC can be expected to attempt to pirate some of our engineers, for both their information and expertise. We need to consider the impact of our reconstruction plans being dissected by our potential major competitor in time for DJC to optimize its response. Given that much of Mr. Li's plan may be implemented in one year, and that it would likely require more than a year for DJC to permit and construct a new plant, there is sufficient time to evaluate the plan with respect to those portions useful on their own merit, and those required as a counter to DJC should it materialize. We now proceed to evaluate the Li Memo by Phase and paragraph, in the light of the above facts. P1.1 – Separate Cell Prototype.
Currently our plant is set up as in Exhibit A, with five process modules and at least 2 inventory holding areas (WIP) in the assembly line. The terminal plating area is bordered by a holding area on each side. Plastic housing manufacture proceeds most swiftly and accumulated WIP has to be delivered to a holding area to await plating of the fabricated electrodes to catch up in order that all components can be delivered to the assembly area. Thus plating capacity is the limiting factor in our assembly process. Mr. Li's proposition that manufacturing could speed up by removing prototype runs to an area across the street ignores the fact that fabricated connectors then must return to the main plant for plating before assembly can continue. We advise that Mr. Li's suggestion could more effectively enhance plant output if additional plating capacity was installed across the street. An analysis should be performed of how that additional plating capacity might speed the entire process. If any item is to be shuffled from one side of the road to another, priority should be given to that item which best augments total output. P1.2 - Quality Control Measures.
If current trends continue, Sunnyvale will produce 510 million units /year in 1996 (85% of design capacity). Given the current defect effect rate of 2.6%, approximately 13,260,000 faulty units would have to be removed by inspectors, or they would be passed along to customers whose products might not then function unless the customer was burdened by the requirement to re-inspect. Mr. Li's goal of improving quality by 50% still leaves a 1.3% failure rate. Statistical Process Control and Design for Experimentation, if implemented to the Six Sigma level, has the capability of reducing error to an average of 3.4 units/million produced. A number of defective units below 2,000 represents a significantly higher yield, and therefore a major difference in quality. We would immediately achieve a 2.5+% in our saleable output. DJC claims to achieve a quality level that produces only 1 defect per million connectors manufactured, or about 510 defects for an entire year of our plant's output, versus about 2,000 for a Six Sigma result or 6.5 million from Mr. Li's proposed 50% improvement. We suggest this contrast reveals a major shortcoming of Mr. Li's proposal. Mr. Li's assumption that employee training in statistical quality control measures combined with automated inspection will reduce defects by only 50% strongly suggests that either the current work force cannot be trained adequately, and/or the contemplated control measures are inadequate. Employees exclaiming that "we do it the old fashioned way" demonstrates an unacceptable and apparently ingrained attitude. Perhaps DJC's...
Please join StudyMode to read the full document