[PDF] Decision rules ISO 3951 is more efficient

View - Download Decision rules

Previous PDF

Next PDF

1LPIS-day, Ispra, April 15th, 2010

Decision rules

2LPIS-day, Ispra, April 15th, 2010

Is this correct?

When is the result to be colored red?rationale

3LPIS-day, Ispra, April 15th, 2010ISO sampling schemes:

•sampling procedure describes:

1.the selection of a sample (or samples) from a lot

2.the inspection or analysis of the sample

3.the classification of the lot (as 'acceptable' or 'not acceptable') based upon

the result of the inspection or analysis of the sample •the precise definition of an acceptance sampling procedure involves the setting or selection of: -the characteristic to be measured -lot size -attribute or variables plan -Limiting Quality (LQ) level, for isolated lots; or the AQL (Acceptable Quality

Limit), for a continuous series of lots

-the level of inspection -the size of the sample -the criteria for acceptance or rejection of the lot -the procedures to be adopted in cases of disputenLevel IIfrom LPISfrom discussion documentoutline

4LPIS-day, Ispra, April 15th, 2010ISO2859 definitions / concepts

Pro Memoria:

Lot: definite quantity of some commodity manufactured or produced under conditions, which are presumed uniform for the purpose

Consignment: quantity of some commodity delivered at one time. It may consist in either a portion of a lot, either a set of several lots.

Sample Set composed of one or several items (or a portion of matter) selected by different means in a population

Item: actual or conventional object on which a set of observations may be made, and which is drawn to form a sample

Representative sample: sample in which the characteristics of the lot from which it is drawn are maintained

Acceptable Quality Level (AQL):the rate of non-conforming items at which a lot will be rejected with a low probability, usually 5 %The AQL is particular producers' risk

Limiting Quality (LQ):the rate of non-conforming items at which a lot will be accepted with a low probability, usually 10 % The LQ is a particular consumers' risk

non-conforming: one or more, quality characteristic does not meet its established quality specification

defect: non-fulfillment of an intended usage requirement.

5LPIS-day, Ispra, April 15th, 2010ISO 2859 for inspection by attributes

•attribute: pass/fail quality characteristic •generally applicable •parts -ISO 2859-0:1995 Part 0: Introduction to the ISO 2859 attribute sampling system -ISO 2859-1:1999/Cor 1:2001 Part 1: Sampling schemes indexed by acceptance quality limit (AQL) for lot-by-lot (continuous production) -ISO 2859-2:1985 Part 2: Sampling plans indexed by limiting quality (LQ) for isolated lot inspection (isolated consignments) -ISO 2859-3:2005 Part 3: Skip-lot sampling procedures (reliable continuous production) -ISO 2859-4:2002 Part 4: Procedures for assessment of declared quality levels (auditing purposes) -ISO 2859-5:2005 Part 5: System of sequential sampling plans indexed by acceptance quality limit (AQL) for lot-by-lot inspection

6LPIS-day, Ispra, April 15th, 2010ISO 3951 for inspection by variables

•variable: measurable (quantitative) quality characteristic •ISO 3951 is more efficient that ISO 2859 (at the same effectiveness) the same 'outcome' with a smaller sample •parts -ISO 3951-1:2005 Part 1: Specification for single sampling plans indexed by acceptance quality limit (AQL) for lot-by-lot inspection for a single quality characteristic and a single AQL -ISO/FDIS 3951-2 Part 2: General specification for single sampling plans indexed by acceptance quality limit (AQL) for lot-by-lot inspection of independent quality characteristics -ISO/DIS 3951-3 Part 3: Double sampling schemes indexed by acceptance quality limit (AQL) for lot-by-lot inspection -ISO 3951-5 Part 5: Sequential sampling plans indexed by acceptance quality limit (AQL) for inspection by variables (known standard deviation)

ISO2859-x and ISO3951-x

•All have the same effectiveness for making a correct or wrong verdict •Each is more efficient for a particular situation (affecting probability conditions)

7LPIS-day, Ispra, April 15th, 2010ETS choices

choice:ISO2859-2: Sampling procedures for inspection by attributes, Part

2: Sampling plans indexed by limiting quality (LQ) for isolated lot

inspection determines sample size n + acceptance number Ac for known (α, β) •Inspection by attributes: pass/fail for each inspected item (parcel) This is not the case for element 1 which is a variable •Limiting quality: the product is expected to be "better",

We express our expectations in terms of LQ

•Isolated lot: no direct feedback to the production process.

LPIS Assumptions

Homogeneous lots: in term of LPIS production and update processes (i.e. lineage and methodology, not necessarily organization, not necessarily land characteristics) Random sampling: in respect of Reference Parcels , not farmer behavior (in practice, CwRS sites with a RP risk component are NOT random)

8LPIS-day, Ispra, April 15th, 2010ISO2859-2 Procedure A

9LPIS-day, Ispra, April 15th, 2010How does it work?

1.our expectation "not more than 1 %"

was indexed on LQ 2% 2

2.Determine sample size n for the lot

1250 for > 500.000

3.compare observed number of non-

conforming items with acceptance No

4.if > 18: lights on

if =< 18; lights off •18/1250 = 1.44% (> 1%) •If your LPIS has 2% nc, it has a 91% chance to be rejected •If your LPIS has 1% nc, it has a 95% chance to be accepted

10LPIS-day, Ispra, April 15th, 2010How does it work (cont'd)?

"not more than 5%" non-conforming parcels indexed on LQ=8

Required sample size is 315, Ac=18

•But you already inspected 1250 (for presence of defects, LQ=2) •315 valid for smaller LPIS sizes

18/315 = 5.714 % equivalent to 71 nc

per 1250 •this is > 5% (or 62 nc)

Rather than take the first 315

from the pre-selection list, "lights on" when more than 71 nc parcels found in sample

11LPIS-day, Ispra, April 15th, 2010practice examples

LPIS non-conformant RP (10202)

(Rate of the Reference Parcels with difference between eligible area observed and eligible area recorded less then 3%) acceptance level (set for LPIS)

5.00% expressed as LQ = 8%

MS X MS Y

pre-selection 2021 3750 sample800 1250 non-conforming 14 67 n/Ac 315/18 315/18

Ac (proportional)4571

Result14<4567<71LPIS potential critical defects

(10205) (Rate of the Reference Parcels with potential defect, which can obstruct the use of the parcel in the LPIS) acceptance level (set for LPIS)

1.00% expressed as LQ = 2%

MS XMS Y

pre-selection 2021 3750 sample800 1250 non-conforming 1 30 n/Ac 800/101250/18

Ac (proportional)n/an/a

Result1<1030>18

12LPIS-day, Ispra, April 15th, 2010expressing non-conforming rates

•Percent non-conforming items Each item (parcel) passes or fails if the non-conforming attribute occurs each parcel counts one e.g. quality element 2 "rate of non-conforming reference parcels..." •Number of non-conformities per 100 items Each item (parcel) is inspected and abundance of non- conformities is counted can count several times e.g. quality element 3 Several causes of the same type in one parcel (e.g. two land changes) Several causes of different type in one parcel (e.g. land change and processing error); effectively 5 seperate "percent non-conforming items" for each cause supported but not applied in 2010

13LPIS-day, Ispra, April 15th, 2010What does LQ mean?

Limiting quality (LQ)

-is the expression of the worst quality lot accepted (consumer's perspective) -Is relevant for a product -not useful for guiding processes: what can you do if you don't accept the quality?

Acceptable quality limit (AQL)

-is an expression for the desired process output (producer's perspective) drives feedback for process improvement, preventing you end up with an unacceptable product If you organise refresh or production processes, monitor the processes through AQL

14LPIS-day, Ispra, April 15th, 2010LQ versus AQL(procedure B!!!!)

•If your LPIS has 2% nc, it has a 91% chance to be rejected •To guide your contractor's production process, to consistently meet the expectation an AQL of 0.65 needs to be targeted0.95

0.650.09

"not more than 1%"

15LPIS-day, Ispra, April 15th, 2010Further reading

ISO standards

-www.iso.org -standards mentioned above -guidance in ISO/TR 8550-1:2007 to 8550-3, -replacing ISO/TR 8550:1994 -ca 50-100 € each

FAO ALINORM 04/27/23 APENDIX III

PROPOSED DRAFT GENERAL GUIDELINES ON SAMPLING

-free of charge -126 page pdf version -not exactly holiday literature ☺

16LPIS-day, Ispra, April 15th, 2010Conclusion

1.our expectation better than 1% indexed on LQ=2%

2.this determines sample size 'n' for all inspections

3.acceptance number ≠proportion in sample (larger!)

4.other LQ and variable indicators are over-sampled +

acceptance number is proportionally adjusted more robust verdict fair trade-off between costs, complication, reliability (risks) use AQL for contractor monitoring (±⅓of LQ) any variation requires a priori agreement

17LPIS-day, Ispra, April 15th, 2010

Thank you for your attention!

quotesdbs_dbs6.pdfusesText_12

[PDF] difference between paper maps and digital maps

[PDF] difference between ppp and traditional procurement

[PDF] différence entre extraction et lavage

[PDF] difference entre langage ecrit et oral

[PDF] différence entre langage évolué et langage machine

[PDF] difference entre langue vernaculaire et dialecte

[PDF] difference langue kabyle et arabe

[PDF] difference of squares trinomial examples

[PDF] different theoretical views on national debt

[PDF] different types of distance measures in clustering

[PDF] digital animation software

[PDF] digital currency 2020 usa

[PDF] digital filter implementation in c

[PDF] digital signature application form for government

[PDF] dijkstra algorithm complexity analysis