Flashcards

BMGT385 - Ch 10

Lauren Roche English 36 cards

Cards in this deck (36)

1when talking about qualityboth mean & variability matter
2statistical process control (SPC)-process variability is inevitable (a process can never produce the same quality always)
-use periodic inspection to check whether process is in control
-management needs to separate natural causes of variation (acceptable & normal) from special causes (ex: machine operator had problems at home & performed worse)
3how much variability is too much?-determine amount of natural variability inherent in system
-determine upper & lower control limits to track process to make sure w/in bounds
4target valuewhat you're shooting for (ex: length of 10 ft for a table)
5tolerance limitsdesign acceptance (ex: length of table can be 10 + or - 1/10 ft)
6cause and effect diagramused to identify & illustrate the main causes of the problem
7check sheetsused to measure the occurrence of each of the various causes summarized in the cause and effect diagram
8pareto chartused to highlight the most important causes
9control charts-track process parameter over time (mean = variable quality characteristic, % defects = attribute)
-distinguish between common (in control) & assignable (out of control) causes of variation
-measure process performance (how much common cause is in the process while it's in control)
-control process performance (is process in control)
10common causes of variability (natural or chance variation)-background noise
-outcome of small unavoidable causes
-"stable system of chance causes"
11assignable causes (special variation)-not part of the chance causes
-identifiable & preventable (hopefully)
12variable quality characteristic control chart-measures a variable on a continuous scale, like inches, mins, etc.
-lots of applications in manufacturing
-examples of service applications: time-to-service (cable customer service), customer service time, X bar & R chart
13attribute data control charts-measures the presence or absence of some key quality characteristic
-success or failure
-p-charts
-examples: on-time or not, defective order or not
14X bar chartplots mean of individual samples
15R chartplots range of individual samples (use w X bar chart to fully interpret process data)
16constructing X bar & R charts
17constructing a p chart1. compute % of defects for each sample & then overall % (p bar) of defects across all samples
2. estimate sample SD & compute control limits
18what x bar, r, & p charts (control) tell1. whether the process is in-control (only natural or common or chance variation is present)
2. whether the process is out-of-control (there are assignable causes of variation in addition)
*aka how consistent the process is (a consistent process doesn't automatically mean high quality...must compare to design specifications)
*tolerance limit is different than control limit
19what x bar, r, & p charts (control) do NOT tell1. what is the assignable cause?
2. is the process meeting the specifications?
20—normal (possibly caused by random variation)
21—lack of stability (possibly caused by assignable causes, like tool, material, operator, overcontrol)
22—cumulative trend (possibly caused by tool wear)
23—cyclical (possibly caused by different work shifts, voltage fluctuations, seasonal effects)
24what happens when we 1st set up the x bar chart & find that 1+ points fall outside the control limits?see if it's an assignable cause or not & bring process into control 1st
25interpreting an X bar & R chart-look at the R chart 1st & if it's out of control, then the control limits on the X bar chart are meaningless
-determine if there are any patterns or trends in the plotted points & if so, then special causes must be eliminated
26commonly accepted guidelines that indicate a process that has gone out of control1. any observation about the UCL or below the LCL
2. 5 points in a row above or below the center line
3. 5 points in a row all increasing or decreasing (indicate the process is not behaving randomly)
4. any other obviously non-random pattern (hugging the control limits, cyclical patterns, etc.)
27do we have to redraw the chart every time the process changes?yes
28do we plot the individual observations of the sample means? what about the range?only plot sample means & sample ranges (some individual points can lie outside the control limits & still the sample mean can be w/in the limits. that's why we use the range.)
29are there applications in service industries?customer complaints, mistakes in invoices or statements (but machine dependent processes are more susceptible to control)
30qualityability of a product/service to consistently meet or exceed customer expectations
31six sigma
32X Sigma (it's hard to reach 6 sigma, so companies will target 4 sigma now)
33common vs assignable causes of variation
34comparing processes: process capability
35process capability measure: Cpmeasures the ability of the process to meet engineering specifications
*if Cp > or = 1, process meets specs 99.7% of the time (at least 3 sigma)
*if Cp > or = 1.33, process is capable
36don't confuse control limits w specification limits...a process can be in control, yet be incapable