Dismiss Notice
You must be a registered member in order to post messages and view/download attached files in this forum.
Click here to register.

Acceptance sampling for Raw Material Receving

Discussion in 'ISO 9001:2015 - Quality Management Systems' started by Manish Sawant, Mar 28, 2016.

  1. Manish Sawant

    Manish Sawant Member

    Joined:
    Mar 28, 2016
    Messages:
    15
    Likes Received:
    1
    Trophy Points:
    2
    Hello, Friends,

    I am looking for your guidance in developing an acceptance sampling plan. I am confused and just want o make sure I make correct decision.

    We are a door assembly and distribution company and we get our raw materials mainly from sole vendors in most cases. I want to make an ASP for inspecting the pallet of Doors Slabs (ASP for attributes) that we receive since the performance of our vendors has not been to the expected standards. The problem is that we receive really small-finite lots of the sizes that is not consistent and vary from as small as 30 to 500. According to Z1.4, for eg., I get sample size of 8 and 32 for 30 and 250 doors respectively with an acceptance # 0 and 1 resp. (AQL= 1.5%). Also, We do not yet have any agreement with the vendors for sampling plans and thus do not know their process capabilities. So, why don't many recommend to have an ASP based on consumers acceptable RQL?

    1. Is it a common practice to develop an ASP assuming AQL without the consent of Vendors?

    2. Or should we form it based on our acceptable RQL?

    3. For such finite sizes, where Hypergeometric distribution is applicable, the sample size increases even more for the same acceptance #. Is it fine to stick with Binomial distribution?

    4. How do we help our vendors use SPC to improve their process capabilities, quality of products?

    5. What is the good criteria to decide if we want to have a Zero acceptance sampling plan?

    I am aware the message is really big to read and so I apologize in advance, but, I just wanted to make sure our exact situation is known.. Please reply. Thanks.

    ------------------------------
    Manish Sawant.
    Graduate Student and Intern.
    Industrial Engineering and Management.
    North Dakota State University
    Fargo, ND.
    7017810405
    ------------------------------
     
  2. QMSmaster

    QMSmaster Active Member

    Joined:
    Dec 29, 2015
    Messages:
    59
    Likes Received:
    9
    Trophy Points:
    7
    I would just use a 0 accept plan. It allows you to have the same protection with smaller samples. Over the last 20 years all of my past employers have abandoned regular incoming acceptance sampling and adopted a 0 Defect policy with suppliers. All would sample inspect the first few shipments and then go dock-to-stock. Supplier quality efforts and expenditures were diverted from incoming inspection to proactive Supplier Development, PPAP, auditing, etc.

    NOTE: I am not an acceptance sampling expert. just an experienced practitioner.
     
  3. Qualmx

    Qualmx Well-Known Member

    Joined:
    Oct 7, 2015
    Messages:
    428
    Likes Received:
    52
    Trophy Points:
    27
    Location:
    Mexico
    Manish
    In iso 9001 , a filter that could be usefil,it is evaluation of suppliers.
    Once you have verified they comply with your needs, you select those in compliance and you may start buying their products with certain confidence,avoiding to make hard inspection to them.

    Additionally you could perform audits to those suppliers pushing them to improve their quality.
    Regards
     
  4. Vthouta

    Vthouta Member

    Joined:
    Dec 20, 2015
    Messages:
    10
    Likes Received:
    2
    Trophy Points:
    2
    We are a door assembly and distribution company and we get our raw materials mainly from sole vendors in most cases. I want to make an ASP for inspecting the pallet of Doors Slabs (ASP for attributes) that we receive since the performance of our vendors has not been to the expected standards. The problem is that we receive really small-finite lots of the sizes that is not consistent and vary from as small as 30 to 500. (The variable sizes doesn't affect the sampling plan)

    According to Z1.4, for eg., I get sample size of 8 and 32 for 30 and 250 doors respectively with an acceptance # 0 and 1 resp. (AQL= 1.5%). Also, We do not yet have any agreement with the vendors for sampling plans and thus do not know their process capabilities. So, why don't many recommend to have an ASP based on consumers acceptable RQL?

    1. Is it a common practice to develop an ASP assuming AQL without the consent of Vendors?
    A: ANSI/ASQ Z1.4 is the classic plan, evolved from MIL-STD-105. This could be used to any commodity codes as part of the decision making tool during receiving inspection.

    2. Or should we form it based on our acceptable RQL?
    A: My suggestion is to follow the industry standard and you could explain the basis of rejection and moving forward how their products are evaluated.

    3. For such finite sizes, where Hypergeometric distribution is applicable, the sample size increases even more for the same acceptance #. Is it fine to stick with Binomial distribution?A: Not sure


    4. How do we help our vendors use SPC to improve their process capabilities, quality of products?
    A: If you could do any of the capability plots on the critical to quality requirements, this will give you an understanding on their capabilities. I suggest you to ask their control plan and investigate the opportunities for improvement.

    5. What is the good criteria to decide if we want to have a Zero acceptance sampling plan?
    A: This depends on several variables. You need to come up how important are at your production facility. need more information to answer. You can switch to c=0 sampling plans, but you would reject the entire lot.

    this material will help you to resolve lot of your queries:
    Link: http://www.ombuenterprises.com/Libr...tance_Sampling_Understanding_How_it_Works.pdf

    Thanks,
    Vik
     
    QMSmaster likes this.
  5. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    470
    Likes Received:
    509
    Trophy Points:
    92
    Location:
    Maine
    The c=0 sampling plan (for AQL based plans) actually provides less protection to both you and your suppliers. This type of plan arose several decades ago when some 'customers' who didn't understand statistics or probability couldn't explain how it was OK to have some defects in your sample and accept the lot even tho they had an Acceptable Quality Level greater than zero. They put the c=0 plan in place to make themselves feel beter but it's like taking valium. you feel great but your quality doesn't get better.

    you can get a copy of the following excellent paper at ASQ through the knowledge center. price is $10 for non members and $5 for members. well worth it.
    Acceptance Sampling? The Enterprise Strikes Back! AS9100c = 0 Plans; when Slogans Supplant Science
     
  6. Bev D

    Bev D Moderator Staff Member

    Joined:
    Jul 30, 2015
    Messages:
    470
    Likes Received:
    509
    Trophy Points:
    92
    Location:
    Maine
     
    Vthouta likes this.
  7. tony s

    tony s Well-Known Member

    Joined:
    Sep 10, 2015
    Messages:
    1,170
    Likes Received:
    866
    Trophy Points:
    112
    Location:
    Laguna Philippines
    If your organization subscribes to ISO/TS 16949, one criteria that you need to consider is on the Clause 7.1.2 which says that "Acceptance criteria shall be defined by the organization and, where required, approved by the customer. For attribute data sampling, the acceptance level shall be zero defects (see 8.2.3.1)".
     

Share This Page