BS 5701-4-2003 Guide to quality control and performance improvement using qualitative (attribute) data - Attribute inspection performance control and improvement《利用质量(品质)数据进行质量控制和性.pdf

《BS 5701-4-2003 Guide to quality control and performance improvement using qualitative (attribute) data - Attribute inspection performance control and improvement《利用质量(品质)数据进行质量控制和性.pdf》由会员分享，可在线阅读，更多相关《BS 5701-4-2003 Guide to quality control and performance improvement using qualitative (attribute) data - Attribute inspection performance control and improvement《利用质量(品质)数据进行质量控制和性.pdf（26页珍藏版）》请在麦多课文档分享上搜索。

1、BRITISH STANDARD BS 5701-4:2003 Guide to quality control and performance improvement using qualitative (attribute) data Part 4: Attribute inspection performance control and improvement ICS 03.120.30 BS 5701-4:2003 This British Standard was published under the authority of the Standards Policy and St

2、rategy Committee on 31 October 2003 BSI 31 October 2003 First published as BS 5701 February 1980 The following BSI references relate to the work on this British Standard: Committee reference SS/4 Draft for comment 02/400887 DC ISBN 0 580 42737 4 Committees responsible for this British Standard The p

3、reparation of this British Standard was entrusted to Technical Committee SS/4, Statistical process control, upon which the following bodies were represented: Association for Road Traffic Safety and Management (ARTSM) BAE Systems British Standards Society (BSS) Clay Pipe Development Association (CPDA

4、) Federation of Small Businesses (FSB) Gauge and Tool Makers Association (GTMA) General Domestic Appliances Ltd. Institute of Quality Assurance National Physical Laboratory Royal Statistical Society Amendments issued since publication Amd. No. Date CommentsBS 5701-4:2003 BSI 31 October 2003 i Conten

5、ts Page Committees responsible Inside front cover Foreword ii 1S c o p e 1 2 Normative references 1 3 Terms, definitions and symbols 1 4 Attribute inspection performance control and improvement 1 5 Inspector error in the detection of discrete events 10 Annex A (normative) Application of signal detec

6、tion theory to the assessment of inspection performance 18 Bibliography 21 Figure 1 Bias and precision 7 Figure 2 Gauge performance profile 10 Figure 3 Role of the value of d in determining the ease of discrimination of a signal 12 Figure 4a) Relationship between criterion response, c, and a hit and

7、 a miss for a given value of discriminatory index, d 13 Figure 4b) Relationship between criterion response, c, and a correct decision and false alarm for a given value of discriminatory index, d 13 Figure 5 Comparison of the results from two observers having identical discriminatory powers but diffe

8、rent decision criterion 14 Figure 6 Operating characteristics showing the effect on performance of changes in the signal discriminatory index, d, through the range of decision criterion, c 15 Figure A.1 Illustration of signal and no signal relationships 19 Table 1 Minimum light levels for different

9、types of inspection 2 Table 2 Example 9 Table 3 Gauge estimated probabilities of acceptance 9 Table 4 Four way classification decision matrix 11 Table 5 Inspector decision matrix 16 Table 6 Results of inspector performance audit 17 Table 7 Characteristics of inspectors 17 Table A.1 Table of ordinate

10、s, y, and p to z conversions for a standardized normal distribution 18BS 5701-4:2003 ii BSI 31 October 2003 Foreword BS 5701-4:2003 partially supersedes BS 5701:1980 and BS 2564:1955 and all four parts of BS 5701 together supersede BS 5701:1980 and BS 2564:1955, which are withdrawn. BS 5701-1, 2 and

11、 3 demonstrated the business benefits, versatility and usefulness of a very simple, yet powerful, pictorial control chart method for monitoring and interpreting qualitative data. This qualitative data can range from overall business figures such as percentage profit to detailed operational data, suc

12、h as percentage absenteeism, individual process parameters and product/service features. The data can either be expressed sequentially in yes/no, good/bad, present/absent, success/failure format, or as summary measures (e.g. counts of events and proportions). For the control charting of measured dat

13、a, refer to BS 5702-1. The treatment of charting of qualitative data in BS 5701-1 is essentially at appreciation level. However, it is intended to provide adequate information for a gainful first application, by a typical less statistically inclined user, in many everyday situations. BS 5701-2 and B

14、S 5701-3 provide a more rigorous, statistical-based, approach to process control and improvement using qualitative data. BS 5701-4 deals with measuring and improving the quality of decision-making in the classification process itself. With the monitoring of measurement data, decisions are made with

15、calibrated instruments against objective criteria. A different situation prevails with attribute data. Decisions are frequently made by uncalibrated people against subjective criteria. Much is made of the need for determining the ongoing capability of measuring instruments. It is standard practice t

16、o have stringent ongoing calibration requirements for instruments. Considerable attention is given to the need for measurement systems analysis in terms of resolution, accuracy and precision to verify, for example, conformance with exacting uncertainty, repeatability and reproducibility criteria. Ye

17、t, surprisingly, relatively little similar thrust takes place with the intrinsically much less reliable subjective decisions taken by people. Consequently, data used to construct control charts to monitor attribute processes frequently lack credibility. A similar situation occurs with acceptance sam

18、pling by attributes. How reliable are the decisions made? How valid are they? The statisticians classify such errors as Type 1 or Type 2. A Type 2 error may be caused by excessive caution, e.g. a jury or court who let a guilty defendant go free, whereas the Type 1 error can result from excessive zea

19、l, e.g. the conviction of an innocent person. Similar conditions prevail in business: faulty items are sometimes accepted and good items are sometimes incorrectly rejected. The situation in business is compounded by the presence of subjective standards against which to make a judgement. In turn, thi

20、s is exacerbated by a number of factors, such as time and commercial pressures, workplace environment and inspector aptitude, faculties, skills and knowledge. BS 5701-4 takes the view that much can, and should, be done to improve decision-making in business when subjective judgements are involved. I

21、t provides guidelines by which this can be achieved. A British Standard does not purport to include all the necessary provisions of a contract. Users of British Standards are responsible for their correct application. Compliance with a British Standard does not of itself confer immunity from legal o

22、bligations. Summary of pages This document comprises a front cover, an inside front cover, pages i and ii, pages 1 to 21 and a back cover. The BSI copyright notice displayed in this document indicates when the document was last issued.BS 5701-4:2003 BSI 31 October 2003 1 1 Scope BS 5701-4 describes

23、methods for enhancing the quality of attribute data used for monitoring, controlling and improving business processes. 2 Normative references The following referenced documents are indispensable for the application of this document. For dated references, only the edition cited applies. For undated r

24、eferences, the latest edition of the referenced document (including any amendments) applies. BS EN ISO 9000:2000, Quality management systems Fundamentals and vocabulary. BS ISO 3534-1, Statistics Vocabulary and symbols Part 1: Probability and general statistical terms. BS ISO 3534-2, Statistics Voca

25、bulary and symbols Part 2: Applied statistics. 3 Terms, definitions and symbols For the purposes of this part of BS 5701, the terms, definitions and symbols given in BS ISO 3534-1, BS ISO 3534-2 and BS EN ISO 9000:2000, Clause 3 apply. 4 Attribute inspection performance control and improvement 4.1 O

26、verview Effective monitoring of discrete data processes is highly dependent on the validity of judgements made on whether a particular attribute, signal or event is present or not. Sometimes the decision is obvious, as objective criteria are present; for example, the number of incoming calls per min

27、ute and whether or not calls are answered within six rings. Frequently, however, subjective judgements are involved, such as the determination whether a particular flaw or imperfection is present or not, or the need to discriminate between acceptable and non-acceptable products. In the case of subje

28、ctive judgements it is particularly important that: a) the decision makers, the assessors, supervisors, inspectors, checkers, viewers, verifiers, auditors, and the like, have the appropriate aptitudes, faculties, knowledge and skills; b) “operational definitions” exist for the particular attributes

29、or events in question; c) the influence of social and organizational factors and work-place design are taken into account; d) assessor calibration and assessment performance audits are conducted where applicable; e) attribute gauges have the required repeatability and accuracy. 4.2 Defining referenc

30、e standards (establishing operational requirements) Discrete data often arises in situations where some form of judgement is applied. To minimize variations in judgement, it is essential that clearly defined conformance criteria are established. These criteria are often expressed in terms of “operat

31、ional definitions” of characteristics. Such an operational definition: a) consists of a criterion to be applied, an examination and a decision: yes or no; b) requires a prescribed examination, a record of the result and a comparison of the result with the agreed criterion; c) should be communicable

32、with the same meaning to one person as to another, yesterday, today and tomorrow. An operational definition should not: i) quote exact values, e.g., round, flat, clean, pucker-free, 20 % fat (minimum, maximum, average?), 40 % wool (how distributed?); ii) use adjectives that have no communicable mean

33、ing e.g. uniform, safe, good or reliable.BS 5701-4:2003 2 BSI 31 October 2003 Written standards for attributes are frequently difficult to develop and to apply without interpretations. In such cases, other methods are used, such as reference samples, photography, line diagrams and stereoscopic pictu

34、res. Often these are graded, e.g. “preferred”, “minimum acceptable” and “unacceptable”, or graded on a scale e.g. 1 to 5. 4.3 Influence of social and organizational factors and work-place design 4.3.1 Social and organizational factors An inspector can be subjected to a number of social pressures fro

35、m those providing him with products to inspect and those receiving them. Sometimes, these pressures balance out. Organizational structures should be such that rational objective criteria prevail. These should always aim to support rather than undermine the technical performance of the inspector. Met

36、hods of payment can also influence inspectors performance. Payment based on total throughput tends to give rise to a tendency to skimp inspection. Payment based on the number of faults found can give rise to a possible increase in commission errors. A further source of potential bias is the reject “

37、norm” for a particular job, or in a particular department. It has been shown that it is possible for these norms to have a greater influence on the proportion of the parts rejected than the actual part quality level. These kinds of pressures can be alleviated by the provision of clear, objective acc

38、eptance standards, adequate instructions and training, effective inspection audit procedure and a sound organizational structure. 4.3.2 Inspection workstation design 4.3.2.1 General Virtually all manual industrial inspection tasks rely heavily on vision, whether one is dealing with surface texture,

39、colour, electrical or mechanical dimensions, correctness of assembly, etc. It is essential, therefore, if one is to ensure effective operating efficiency in terms of speed and accuracy, to take account of the significant factors affecting visual performance. 4.3.2.2 Lighting 4.3.2.2.1 Aspects of lig

40、hting Aspects of lighting of particular relevance to visual inspection are: a) light level; b) luminance; c) contrast; d) colour; e) glare. 4.3.2.2.2 Light level The rate of flow of luminous energy, (namely light) from a lamp is given in terms of lumens (lm). For example, the light output from: a) A

41、 100 W tungsten filament bulb is 1 200 lumens; b) An 80 W fluorescent lamp is 4 000 lumens. Not all the light output from a lamp can be transferred to the viewing plane. The illumination of a work surface is expressed in terms of “lumens per unit area”. Lumens per square metre is termed “lux”. The o

42、rder of service illumination proposed for different classes of inspection work is shown in Table 1. Table 1 Minimum light levels for different types of inspection Type of inspection Minimum service value of illumination lux Minute detail (e.g. watches) 2 000 to 3 000 Small detail (e.g. gauging) 1 50

43、0 to 2 000 General viewing 500 to 1 000BS 5701-4:2003 BSI 31 October 2003 3 4.3.2.2.3 Luminance Luminance is the light emitted or reflected to the eye from an object, as opposed to illumination, which is the light falling onto it. Luminance is what is seen when inspecting an object, not illumination

44、. The brightness (or luminance) of the object is important when considering the amount of light required for a particular viewing task. This luminance depends both on the level of illumination of the object and the proportion of that incident light which is reflected from the surface of the object.

45、Two situations arise: a) non-specular surface a perfectly diffusing surface, e.g. blotting paper that has the same luminance from all angles of view. Here: Luminance = Illumination Reflecting ratio B = ER This relationship is often used for surfaces that are not perfect diffusers: b) specular surfac

46、es bright polished surface (e.g. mirror, bright metal). Here, the luminosity will depend not only on the proportion of light reflected by the surface itself but by the image of a concentrated light source itself seen reflected in the surface. This luminance will have a directional property depending

47、 on the relative positions of the light source, observer and object being viewed. Here, luminance is expressed as intensity in a specified direction/projected area as seen from that direction. It should be appreciated, however, that specular surfaces also have reflectivity ratios in the same way as

48、perfect diffusers. However, they introduce an additional complication, e.g. if one is inspecting sheets of polished aluminium, the light source itself or surroundings can be reflected in the surface at certain angles of view, thus impairing viewing efficiency. 4.3.2.2.4 Contrast Regardless of the le

49、vel of illumination, an object can only be seen by virtue of contrast with its background. There are two forms of contrast: brightness and colour. It is possible for two quite distinct colours to reflect the same amount of incident light brightness contrast yet be clearly distinguishable purely by their difference in hue. 4.3.2.2.5 Colour A sensation of light in the human eye is caused by electromagnetic radiation within the wavelength band 400 nm to 75