The assignable cause The Control Chart Statistical basis of the control chart Control limits, false and true alarms and the operating characteristic function
Lecture 10: Introduction to Statistical Process Control
1
Spanos
EE290H F05
Managing Variation over Time • Statistical Process Control often takes the form of a continuous Hypothesis testing. • The idea is to detect, as quickly as possible, a significant departure from the norm. • A significant change is often attributed to what is known as an assignable cause. • An assignable cause is something that can be discovered and corrected at the machine level.
Lecture 10: Introduction to Statistical Process Control
2
Spanos
EE290H F05
What is the Assignable Cause? • An "Assignable Cause" relates to relatively strong changes, outside the random pattern of the process. • It is "Assignable", i.e. it can be discovered and corrected at the machine level. • Although the detection of an assignable cause can be automated, its identification and correction often requires intimate understanding of the manufacturing process. • For example... – – – –
Lecture 10: Introduction to Statistical Process Control
3
Spanos
EE290H F05
Example: Investigate furnace temp and set up a real-time alarm. 3 2 1 0 -1 -2 -3
0
5000
10000
15000
20000
time
The pattern is obvious. How can we automate the alarm? Lecture 10: Introduction to Statistical Process Control
4
Spanos
EE290H F05
The purpose of SPC A. Detect the presence of an assignable cause fast. 2. Minimize needles adjustment. • Like Hypothesis testing – (A) means having low probability of type II error and – (B) means having low probability of type I error.
• SPC needs a probabilistic model in order to describe the process in question.
Lecture 10: Introduction to Statistical Process Control
5
Spanos
EE290H F05
Example: Furnace temp differential (cont.) Group points and use the average in order to plot a known (normal) statistic. Assume that the first 10 groups of 4 are in Statistical Control. Limits are set for type I error at 0.05. 2 UCL 1.2
1 0 -1 -2
LCL -1.2 0
Lecture 10: Introduction to Statistical Process Control
10
20
30 6
Spanos
EE290H F05
Example (cont.) • The idea is that the average is normally distributed. • Its standard deviation is estimated at .6333 from the first 10 groups. • The true mean (μ) is assumed to be 0.00 (furnace temperature in control). • There is only 5% chance that the average will plot outside the μ+/- 1.96 σ limits if the process is in control. In general: UCL = μ + k σ LCL = μ - k σ where μ and σ relate to the statistic we plot. Lecture 10: Introduction to Statistical Process Control
7
Spanos
EE290H F05
Another Example Original data Plot
Averaged Data (n=5) Variable Control Charts 1.070
1.100
1.050
1.050
1.030
0.990 0.970 0.950 LCL=0.9387 0.930
0.850 0
100
200
300
400
500
100
0.900
µ0=1.0006
75
0.950
1.010
50
1.000
UCL=1.0626
25
Mean of small shift
small shift
1.150
small shift
Mean of small shift Lecture 10: Introduction to Statistical Process Control
8
Spanos
EE290H F05
How the Grouping Helps Small Group Size, large β.
Large Group Size, smaller β for same α.
Bad Good
Lecture 10: Introduction to Statistical Process Control
9
Spanos
EE290H F05
Average Run Length • If the type I error (α) depends on the original (proper) parameter distribution and the control limits, ... • ... the type II error (β) depends on the position of the shifted (faulty) distribution with respect to the control limits. • The average run length (ARL) of the chart is defined as the average number of samples between alarms. • ARL, in general, is 1/α when the process is good and 1/(1-β) when the process is bad. Lecture 10: Introduction to Statistical Process Control
10
Spanos
EE290H F05
The Operating Characteristic Curve The Operating Characteristic of the chart shows the probability of missing an alarm vs. the actual process shift. Its shape depends on the statistic, the subgroup size and the control limits. β These curves are drawn Fig. 4-5 from Montgomery, pp. 0.05 110 for α =
deviation in #σ Lecture 10: Introduction to Statistical Process Control
11
Spanos
EE290H F05
Pattern Analysis 3 2 1 0 -1 -2 -3
0
20
40
60
80
100
Other rules exist: Western Electric, curve fitting, Fourier analysis, pattern recognition... Lecture 10: Introduction to Statistical Process Control
12
Spanos
EE290H F05
Example: Photoresist Coating • During each shift, five wafers are coated with photoresist and soft-baked. Resist thickness is measured at the center of each wafer. Is the process in control? • Questions that can be asked: a) Is group variance "in control"? b) Is group average "in control"? c) Is there any difference between shifts A and B? • In general, we can group data in many different ways.
Lecture 10: Introduction to Statistical Process Control
13
Spanos
EE290H F05
Range and x chart for all wafer groups. 600
UCL 507.09
500 400 300 200 100 0 8000
LCL 0.0 UCL 7971.32
7900 7800 7700 7600
LCL 7694.52 0
Lecture 10: Introduction to Statistical Process Control
10
20
30
40 Wafer Groups
14
Spanos
EE290H F05
Comparing runs A and B Range, Shift A
Range, Shift B
600
600
550
500
465 400
500
400
300
300
260
220
200
200
100
100
0 0
10
0
20
0
Mean, Shift A 8000
10
Mean, Shift B
20
8000
7985 7958 7900
7900
7835
7831 7800
7800
7704
7700
7700
7685
7600
7600 0
10
Lecture 10: Introduction to Statistical Process Control
20
0
10
20
15
Spanos
EE290H F05
Why Use a Control Chart? • Reduce scrap and re-work by the systematic elimination of assignable causes. • Prevent unnecessary adjustments. • Provide diagnostic information from the shape of the non random patterns. • Find out what the process can do. • Provide immediate visual feedback. • Decide whether a process is production worthy.
Lecture 10: Introduction to Statistical Process Control
16
EE290H F05
Spanos
The Control Chart for Controlling Dice Production
Lecture 10: Introduction to Statistical Process Control
17
Spanos
EE290H F05
The Reference Distribution 7 6 5 4 3 2 1 2
3
Lecture 10: Introduction to Statistical Process Control
4
5
6
7
8
9 10 11 12 18
Spanos
EE290H F05
The Actual Histogram 7 6 5 4 3 2 1 2
3
Lecture 10: Introduction to Statistical Process Control
4
5
6
7
8
9 10 11 12 19
Spanos
EE290H F05
In Summary • To apply SPC we need: • Something to measure, that relates to product/process quality. • Samples from a baseline operation. • A statistical “model” of the variation of the process/product. • Some physical understanding of what the process/product is doing.
Lecture 10: Introduction to Statistical Process Control