Quality Control : This example shows images with scrolling text. The
controller and borders are also hidden.
The JavaScript code used to launch this exposition is:
new Scriptio({script:'qcproc.txt', soundtrack:'qcproc.swf',
width:592, height:418});
The audio soundtrack is done using Flash audio. The total audio file size is 300KB, and it does not
have to be fully-downloaded before the soundtrack starts playing.
script: qcproc.txt
:set title to "Using A Control Chart To Monitor A Process" :set background to cream +.5 :display "Control Charts" at {40,40} size 16 style bold color black face "verdana" label "cont" fade 1.0 :picture "standard.gif" at {280,215} size {268,163} fade 1 label "standard" Control charts help us +2.9 :display "Find unwanted variation." under "cont" by {0,20} bullet true color black size 12 style plain fade 1 label "bull1" +3.7 :picture "top.gif" at "standard" size {268,84} wipe 2 direction up label "top" :picture "bottom.gif" under "top" by {0,-1} size {268,80} wipe 2 direction down label "bottom" find unwanted variation and +5.1 :display "Ensure process is in a state of " under "bull1" by {0,10} bullet true color black size 12 style plain fade 1 label "bull2" +6.7 :display "control" after "bull2" color dkred fade 1 size 12 label "ensures2" +7 :clear "top" wipe 2 direction down :clear "bottom" wipe 2 direction up ensure that our process is in a state of statistical control. It is important to understand, however, that statistical control simply means that the process is +14.1 :clear "ensures2" +14.4 :display "stability" after "bull2" fade .5 label "stable" stable. In other words, +16.9 :clear "stable" +17.1 :display "control" after "bull2" fade .5 color dkred label "ensure2" +17.6 :display ", that is," after "ensure2" by {0,0} color black fade 1 size 12 label "note1" :display "the variation in the process is just random variation." under "bull2" by {12,4} fade 1 size 12 label "note2" that the variation observed in the process is what we would expect from just random variation. The process might very well be +24.4 :line under "standard" by {0,-47} color dkblue width 2 fade 1 size {266,0} label "line" :display "Target" at "line" by {-45,-6} fade 1.5 label "ds" +25.4 :display "Statistical control does not refer to the process being on target" under "note2" by {-12,10} bullet true color black size 12 style plain fade 1 label "bull3" :display "or having acceptable level of variation." under "bull3" by {12,4} fade 1 label "donot2" off target or have a larger variation than desired. Process capability (a topic discussed in another section) +33.6 :clear {"line","ds"} fade 1 :display "Capability analysis does this." after "donot2" by {9,0} fade 1 label "donot3" addresses how a process performs relative to target and specification. But we cannot assess process capability unless our process is in statistical control. Otherwise what we observe now may not reflect what will happen in the future.<br/> +49 :display "" at "standard" by {-5,-10} label "a" :clear "standard" +49.4 :picture "rchart.gif" at "a" size {255,198} fade .5 label "rchart" :display "There are many types of" under "donot2" by {-12,10} bullet true fade 1 label "bull4" :display "control charts, " under "bull4" by {12,4} fade 1 label "cc" +52.5 :clear "rchart" +53 :picture "uchart.gif" at "a" size {279,196} fade .5 label "uchart" +56.5 :clear "uchart" +57 :picture "pchart.gif" at "a" size {278,195} fade .5 label "pchart" +61.5 :clear "pchart" Many different types of control charts are in use today. We'll illustrate a few common ones in the next set of activities. Though the particular quality characteristic and nature of the data determine which chart is appropriate, +62 :picture "imrchart.gif" at "a" size {272,201} fade .5 label "imchart" +62.8 :display "but almost all" after "cc" fade 1 label "almost" :display "share the same basic principal." under "cc" by {0,4} fade 1 label "share" almost all control charts share the same basic principle. +67.4 :pause +68, clear :clear Samples from a process are selected over time. For each sample, we compute a statistic such as the mean, the range, +69.1 :picture "machine_l.gif" at {15,240} size {100,138} label "machine_l" :picture "machine_r.gif" after "machine_l" size {232,138} label "machine_r" :picture "fade.png" after "machine_r" by {-60,0} size {97,152} label "fade" :picture "tray2.png" at "machine_l" by {85,79} size {82,27} label "tray2" :picture "rings2.png" at "tray2" by {7,7} size {65,14} label "rings2" :picture "tray1.png" at "machine_l" by {160,79} size{82,27} label "tray1" :picture "rings1.png" at "tray1" by {16,3} size {60,17} label "rings1" :picture "tray.png" at "machine_l" by {235,79} size {82,27} label "tray" :picture "tray.png" at "machine_l" by {-365,79} size {82,27} label "last" :picture "tray3.png" at "machine_l" by {10,79} size {82,27} label "tray3" :picture "rings3.png" at "tray3" by {47,3} size {21,17} label "rings3" :layer {"tray3","rings3","last"} :layer "machine_l" :layer "fade" +69.5 :slide {"tray","tray1","tray2","tray3","rings1","rings2","rings3","last"} by {35,0} duration 1.5 +70.8 :slide "rings1" to {50,50} duration 1 :clear "rings1" grow 1.0 +71.7 :picture "big3.gif" at {40,40} size {151,70} grow .5 +72.2 :slide {"tray","tray1","tray2","tray3","rings1","rings2","rings3","last"} by {35,0} duration 3.5 :move "last" by {300,0} :clear "tray" wipe 1.5 direction left +74 :display "Mean" at {370,160} size 12 color black style plain label "mean" fade .5 +75.3 :display "Range" under "mean" by {0,3} label "range" fade .5 :slide "rings2" to {210,50} duration 1 :clear "rings2" grow 1.0 +76.3 :display "Standard Deviation" under "range" by {0,3} label "std" fade .5 :picture "big3.gif" at {200,40} size {151,70} grow .5 the standard deviation, or the +77.3 :display "# of Defective Parts" under "std" by {0,3} label "parts" fade .5 -:picture "big3.gif" at {200,40} grow .5 +77.8 :slide {"tray1","tray2","tray3","rings1","rings2","rings3","last"} by {35,0} duration 1.5 -:clear "tray1" wipe 1.5 direction left +79 :slide "rings3" to {375,50} duration 1 :clear "rings3" grow 1.0 count of the number of defective parts. This statistic is +80 :picture "big3.gif" at {365,40} size {151,70} grow .5 :picture "graph.gif" at {354,270} size {189,102} fade 1.0 label "graph1" :layer "parts" then plotted over time.<P> +82.5 :slide {"tray1","tray2","tray3","rings1","rings2","rings3","last"} by {35,0} duration 1.5 :clear "tray1" wipe 1.5 direction left :clear except "graph1" +83 :slide "graph1" by {-150,-130} duration .5 +83.5 -:picture "graph2" at ((location "graph1")-((extent "graph1")/2)) grow 1.0 label "graph2" :picture "graph2.gif" at {Left("graph1")-(Width("graph1")/2), Top("graph1")-(Height("graph1")/2)} size {378,204} grow 1.0 label "graph2" :layer "graph2" after "graph1" +84 :clear "graph1" ~1000 +85 :line at "graph2" by {1,104} size {375,0} color green width 2 wipe 1 direction right label "center" :line at "graph2" by {1,24} size {375,0} color red width 2 wipe 1 direction right label "control1" :line at "graph2" by {1,179} size {375,0} color red width 2 wipe 1 direction right label "control2" We add a center line and control limits to the graph so that we can easily see when the statistic becomes unusual. The details of finding these lines are the subject of future lessons.<P> If the process is in statistical control, then +99.3 :picture "hilitgraph.gif" at "graph2" size {378,204} wipe 1 direction right label "hilite" :line at "graph2" by {1,104} size {375,0} color green width 2 wipe 1 direction right label "center2" :line at "graph2" by {1,24} size {375,0} color red width 2 wipe 1 direction right label "control5" :line at "graph2" by {1,179} size {375,0} color red width 2 wipe 1 direction right label "control6" almost all the points on the graph will be within the control limits. +103.3 :clear "hilite" Points +104.9 :picture "hilitgraph2.gif" at "graph2" size {378,204} wipe 1 direction right label "hilite2" :line at "graph2" by {1,104} size {375,0} color green width 2 wipe 1 direction right label "center1" :line at "graph2" by {1,24} size {375,0} color red width 2 wipe 1 direction right label "control3" :line at "graph2" by {1,179} size {375,0} color red width 2 wipe 1 direction right label "control4" outside the control limits indicate that the process was probably out of control during that +111.1 :clear "hilite2" time. We also look for +112.6 :picture "dw.gif" at "graph2" by {0,0} size {378,204} label "dw" :layer {"center","control1","control2"} any clear +115.7 :picture "uw.gif" at "graph2" by {0,0} size {378,204} label "uw" :layer {"center1","control3","control4","plh"} +117 :clear "dw" patterns in the graph. Such patterns can also indicate that the process is out of control. The next activity illustrates some of these patterns. <P> +124.8 :clear "uw" Some people use the terms +126.1 :display "Common cause variation: " under "graph2" by {-58,30} style bold size 12 color black wipe .5 direction right label "ccv1" common cause variation and +128.5 :display "Assignable cause variation: " under "ccv1" by {0,15} style bold wipe .5 direction right label "acv1" assignable cause variation. Common cause variation is variation that is +133.7 :display "random and naturally inherent in the process" after "ccv1" by {0,0} color black size 12 wipe .5 style plain direction right label "ccv2" random and naturally inherent in the process. Assignable cause variation is variation +140.1 :display "caused by change in process" after "acv1" by {0,0} wipe .5 direction right label "acv2" caused by something unusual, such as equipment problems, or an incorrect adjustment, or changes in raw material. +147.7 :pause "Summary" +147.8, clear :clear :picture "summary.gif" size {592,59} with center at 357 label "summpic" :display "Summary" at {40,40} size 12 color dkblue style bold label "summary" :display "Control charts help us determine if a process is in statistcal control." with bullet under "summary" by {0,14} size 12 style plain color black label "summ1" :display "Statistical control means the variability in the process is just random " with bullet under "summ1" by {0,10} label "summ2" :display "variation." under "summ2" by {12,0} label "summ2a" :display "A process that is in statistical control can still be off target or posess " with bullet under "summ2a" by {-10,10} label "summ3" :display "more variation than desired." under "summ3" by {12,0} label "summ3a" :display "We cannot assess a process, for example check to see if it is on " with bullet under "summ3a" by {-8,10} label "summ4" :display "target, unless the process is in statistical control." under "summ4" by {12,0} label "summ4a" :display "Control charts are generally constructed in the same way." with bullet under "summ4a" by {-12,10} label "summ5" :display "Samples from a process are selected over time." with bullet under "summ5" by {25,5} size 12 label "summ5a" :display "For each sample, a statistic such as the mean or range is computed." with bullet under "summ5a" by {0,0} label "summ5b" :display "This statistic is then plotted over time." with bullet under "summ5b" by {0,0} label "summ5c" :display "Any unusual patterns in a control chart can indicate the process is out " with bullet under "summ5c" by {-25,10} size 12 label "summ6" :display "of control." under "summ6" by {12,0} label "summ6a"