|
     
- UID
- 21662
- 积分
- 76093
- 积极性
- 178
- 来自
- 上海
- 在线时间
- 4465 小时
- 注册时间
- 2007-10-14
|
22#
发表于 2009-3-14 14:45
| 只看该作者
本帖最后由 hnwang 于 2009-3-14 14:48 编辑
有志研究研究的人,可以从以下得到很多资料和信息。
Posted by: Jsev607
Posted on: Wednesday, 23rd April 2008, 10:15 PM.
Option 1: Continued Acceptance Sampling
Take a look at: http://www.variation.com/library.html . Once you've read and understood the articles you can either a.) Get yourself a binomial table, determine what risks your customers are willing to accept within one lot (your beta risk) and develop sampling tables accordingly. or b.) get yourself a canned set of c=0 acceptance plans (if you intend to stick with c=0 that is). You could try: MIL-STD-1916 for this.
Option 2: Cease Dependency on Inspections & Develop a Robust Process Control System
Ultimately, inspections are a non-value added activity. We all know you're not going to inspect quality into a product. Map out your process and perform a PFMEA. Lower your associated risks by applying principles such as poka-yoke, DoE, SPC on key inputs, Preventative Maintenance, etc.. Develop a robust process and you can pretty much cease inspection altogether or at least justify dropping sample sizes down to whatever level you feel is manageable. MIL-STD-1916 also helps give you guidelines for this as well though you will need quite a bit of supplementary information to make it a reality.
Good luck. |
|