Statistical process control

From Wikipedia, the free encyclopedia

Jump to: navigation, search
Statistical process control (SPC) is the application of statistical methods to the
monitoring and control of a process to ensure that it operates at its full potential to
produce conforming product. Under SPC, a process ehaves predictaly to produce as
much conforming product as possile !ith the least possile !aste. While SPC has een
applied most fre"uently to controlling manufacturing lines, it applies e"ually !ell to any
process !ith a measurale output. #ey tools in SPC are control charts, a focus on
continuous improvement and designed e$periments.
%uch of the po!er of SPC lies in the aility to e$amine a process and the sources of
variation in that process using tools that give !eight to o&ective analysis over su&ective
opinions and that allo! the strength of each source to e determined numerically.
'ariations in the process that may affect the "uality of the end product or service can e
detected and corrected, thus reducing !aste as !ell as the likelihood that prolems !ill
e passed on to the customer. With its emphasis on early detection and prevention of
prolems, SPC has a distinct advantage over other "uality methods, such as inspection,
that apply resources to detecting and correcting prolems after they have occurred.
(n addition to reducing !aste, SPC can lead to a reduction in the time re"uired to produce
the product or service from end to end. )his is partially due to a diminished likelihood
that the final product !ill have to e re!orked, ut it may also result from using SPC data
to identify ottlenecks, !ait times, and other sources of delays !ithin the process.
Process cycle time reductions coupled !ith improvements in yield have made SPC a
valuale tool from oth a cost reduction and a customer satisfaction standpoint.
Contents
*hide+
, -istory
. /eneral
0 -o! to Use SPC
1 2eferences
3 4iliography
5 See also
6 7$ternal links
[edit] History
Statistical process control !as pioneered y Walter 8. She!hart in the early ,9.:s. W.
7d!ards ;eming later applied SPC methods in the United States during World War ((,
therey successfully improving "uality in the manufacture of munitions and other
strategically important products. ;eming !as also instrumental in introducing SPC
methods to Japanese industry after the !ar had ended.
*,+*.+
She!hart created the asis for the control chart and the concept of a state of statistical
control y carefully designed e$periments. While ;r. She!hart dre! from pure
mathematical statistical theories, he understood that data from physical processes seldom
produces a <normal distriution curve< =a /aussian distriution, also commonly referred
to as a <ell curve<>. -e discovered that oserved variation in manufacturing data did not
al!ays ehave the same !ay as data in nature =for e$ample, 4ro!nian motion of
particles>. ;r. She!hart concluded that !hile every process displays variation, some
processes display controlled variation that is natural to the process =common causes of
variation>, !hile others display uncontrolled variation that is not present in the process
causal system at all times =special causes of variation>.
*0+
(n ,9?9, the Soft!are 7ngineering (nstitute introduced the notion that SPC can e
usefully applied to non@manufacturing processes, such as soft!are engineering processes,
in the Capaility %aturity %odel =C%%>. )his idea e$ists today !ithin the Aevel 1 and
Aevel 3 practices of the Capaility %aturity %odel (ntegration =C%%(>. )his notion that
SPC is a useful tool !hen applied to non@repetitive, kno!ledge@intensive processes such
as engineering processes has encountered much skepticism, and remains controversial
today.
*1+*3+*5+
[edit] General
)he follo!ing description relates to manufacturing rather than to the service industry,
although the principles of SPC can e successfully applied to either. For a description and
e$ample of ho! SPC applies to a service environment, refer to 2oerts =.::3>.
*6+
SPC has
also een successfully applied to detecting changes in organiBational ehavior !ith Social
Cet!ork Change ;etection introduced y %cCulloh =.::6>. Selden descries ho! to use
SPC in the fields of sales, marketing, and customer service, using ;emingDs famous 2ed
4ead 7$periment as an easy to follo! demonstration
*?+
.
(n mass@manufacturing, the "uality of the finished article !as traditionally achieved
through post@manufacturing inspection of the productE accepting or re&ecting each article
=or samples from a production lot> ased on ho! !ell it met its design specifications. (n
contrast, Statistical Process Control uses statistical tools to oserve the performance of
the production process in order to predict significant deviations that may later result in
re&ected product.
)!o kinds of variation occur in all manufacturing processes: oth these types of process
variation cause suse"uent variation in the final product. )he first is kno!n as natural or
common cause variation and consists of the variation inherent in the process as it is
designed. Common cause variation may include variations in temperature, properties of
ra! materials, strength of an electrical current etc. )he second kind of variation is kno!n
as special cause variation, or assignale@cause variation, and happens less fre"uently than
the first. With sufficient investigation, a specific cause, such as anormal ra! material or
incorrect set@up parameters, can e found for special cause variations.
For e$ample, a reakfast cereal packaging line may e designed to fill each cereal o$
!ith 3:: grams of product, ut some o$es !ill have slightly more than 3:: grams, and
some !ill have slightly less, in accordance !ith a distriution of net !eights. (f the
production process, its inputs, or its environment changes =for e$ample, the machines
doing the manufacture egin to !ear> this distriution can change. For e$ample, as its
cams and pulleys !ear out, the cereal filling machine may start putting more cereal into
each o$ than specified. (f this change is allo!ed to continue unchecked, more and more
product !ill e produced that fall outside the tolerances of the manufacturer or consumer,
resulting in !aste. While in this case, the !aste is in the form of <free< product for the
consumer, typically !aste consists of re!ork or scrap.
4y oserving at the right time !hat happened in the process that led to a change, the
"uality engineer or any memer of the team responsile for the production line can
trouleshoot the root cause of the variation that has crept in to the process and correct the
prolem.
SPC indicates !hen an action should e taken in a process, ut it also indicates !hen CF
action should e taken. 8n e$ample is a person !ho !ould like to maintain a constant
ody !eight and takes !eight measurements !eekly. 8 person !ho does not understand
SPC concepts might start dieting every time his or her !eight increased, or eat more
every time his or her !eight decreased. )his type of action could e harmful and possily
generate even more variation in ody !eight. SPC !ould account for normal !eight
variation and etter indicate !hen the person is in fact gaining or losing !eight.
[edit] How to Use SPC
Statistical Process Control may e roadly roken do!n into three sets of activities:
understanding the processE understanding the causes of variationE and elimination of the
sources of special cause variation.
(n understanding a process, the process is typically mapped out and the process is
monitored using control charts. Control charts are used to identify variation that may e
e due to special causes, and to free the user from concern over variation due to common
causes. 4y the nature of the control chart, <understanding the process< is a continuous
activity. With a stale process that does not trigger any of the detection rules for a control
chart, a process capaility analysis is also performed to evaluate the aility of the current
process to produce conforming =i.e. !ithin specification> product.
When, through the control charts, variation that is due to special causes is identified, or
the process capaility is found lacking, additional effort is e$erted to determine causes of
that variance and eliminate it. )he tools used include (shika!a diagrams, designed
e$periments and Pareto charts. ;esigned e$periments are critical to this phase of SPC, as
they are the only means of o&ectively "uantifying the relative importance of the many
potential causes of variation.
Fnce the causes of variation have een "uantified, effort is spent in eliminating those
causes that are oth statistically and practically significant =i.e. a cause that has a only
small ut statistically significant effect may not e considered cost@effective to fi$E
conversely, a cause that is not statistically significant cannot e considered practically
significant>. /enerally, this includes development of standard !ork, error@proofing and
training. 8dditional measures may e re"uired, especially if there is a prolem !ith
process capaility.
Quality control
From Wikipedia, the free encyclopedia
Jump to: navigation, search
%aintenance check of electronic e"uipment on a U.S. Cavy aircraft.
G@ray Boom series of a net!ork adapter card.
Quality control is a process y !hich entities revie! the "uality of all factors involved in
production. )his approach places an emphasis on three aspects:
,. 7lements such as controls, &o management, defined and !ell managed
processes
*,+*.+
, performance and integrity criteria, and identification of records
.. Competence, such as kno!ledge, skills, e$perience, and "ualifications
0. Soft elements, such as personnel integrity, confidence, organiBational culture,
motivation, team spirit, and "uality relationships.
)he "uality of the outputs is at risk if any of these three aspects is deficient in any !ay.
Contents
*hide+
, )otal "uality control
. (ndustrial resources
0 Scientific resources
1 8cademic resources
3 See also
5 Cotes and references
6 Further reading
[edit] Total quality control
)otal Huality Control is the most important inspection control of all in cases !here,
despite statistical "uality control techni"ues or "uality improvements implemented, sales
decrease.
(f the original specification does not reflect the correct "uality re"uirements, "uality
cannot e inspected or manufactured into the product.
For instance, the parameters for a pressure vessel should include not only the material
and dimensions, ut also operating, environmental, safety, reliaility and maintainaility
re"uirements.