2021 INDU372 Final
2021 INDU372 Final
2021 INDU372 Final
Instruction:
In each of the following multiple choice questions, select and write down the correct answers. For example, if you
believe that for Question 1, A, B, C are correct, write down your answer in the following way:
1: A, B, C
If indeed A, B, C are correct, you will receive 3 marks (1 mark for each answer). If your answer is A, B, C, D, but
D is wrong, 1 mark will be deducted and you will receive 2 marks for Question 1. For Problem 1, the maximum
total mark is 15 and the minimum total mark is 0.
Note: There should be 1 or more correct answers in each of the multiple choice questions.
A) It may rely on statistical process control since SPC is the most effective tool for quality improvement
B) It may rely on sampling inspection following industrial standard as this procedure is the most effective
tool for quality improvement
C) It requires coordinated efforts of the company using quantitative, qualitative and managerial tools for
quality improvement
D) It must use tightened inspection to monitor production processes
E) It should have a complete program that involves statistical tools, management commitment, workforce
training, marketing, cost analysis
2. A process with σ =1.0 was in control and an x chart was set. A shift of the process mean took place due to
assignable causes. Assuming that the new process mean is at the lower control limit of the original x control
chart. In using the same x chart to monitor the shifted process,
3. To properly conduct the acceptance sampling inspection, the unit(s) of the samples to be inspected
1
4. For different probability distributions often used for statistical quality control,
6. Assume that 𝑝𝑝1 is acceptable quality level (AQL), 𝑝𝑝2 is lot tolerance percent defective (LTPD), 𝛼𝛼 is producer’s
risk, 𝛽𝛽 is consumer’s risk, N is the lot size. To determine the sample size n and the acceptance number c of a
single sampling plan using a Binomial nomograph, the following specific values of parameters are used:
A) (𝑝𝑝1 , 1 − 𝛼𝛼) and N
B) (𝑝𝑝2 , 𝛽𝛽) and N
C) (𝑝𝑝1 , 1 − 𝛼𝛼) and (𝑝𝑝2 , 𝛽𝛽)
D) (𝑝𝑝1 , 1 − 𝛼𝛼), (𝑝𝑝2 , 𝛽𝛽) and N
E) (𝑝𝑝1 , 𝑝𝑝2 ) and N
7. Military standard 105E (ANSI/ASQC Z1.4, ISO 2859) sampling procedures for inspection by attributes
8. In reliability analysis, a system is composed of 3 identical components with 2 out of the 3 units redundant for
backup purpose. Let 𝑅𝑅𝑠𝑠 be the system reliability and 𝑅𝑅𝑐𝑐 be the component reliability. Then:
A) 𝑅𝑅𝑠𝑠 = 3 × 𝑅𝑅𝑐𝑐
1
B) 𝑅𝑅𝑠𝑠 = × 𝑅𝑅𝑐𝑐
3
C) 𝑅𝑅𝑠𝑠 = 𝑅𝑅𝑐𝑐 × 𝑅𝑅𝑐𝑐 × 𝑅𝑅𝑐𝑐 = 𝑅𝑅𝑐𝑐3
D) 𝑅𝑅𝑠𝑠 = 1 − 𝑅𝑅𝑐𝑐3
E) 𝑅𝑅𝑠𝑠 = 1 − (1 − 𝑅𝑅𝑐𝑐 )3
9. In reliability analysis, a series system is composed of 3 identical components and the system will fail if any
one of the 3 components fails. Let 𝜆𝜆𝑠𝑠 be the system failure rate and 𝜆𝜆𝑐𝑐 be the component failure rate. Then:
2
Problem 2 (16 Marks)
The manager of a wood processing shop wants to use SPC charts to monitor the length of the wood beams the
shop produces using an automatic wood saw. The product specifications are at 1000 ± 3mm. 15 samples are
randomly taken as shown in the following table. It is known that the beam length follows Normal distribution.
Production rate is 1,500 units/hour.
7) It is calculated that the sample standard deviation of the 60 measured beam lengths (from the same table
given below) is 𝑠𝑠 = 1.31. Calculate process capability ratio 𝐶𝐶𝑝𝑝 based on this sample standard deviation.
8) Briefly explain the advantages and possible limitations of conducting process capability analysis without
using control charts comparing to that using control charts.
3
Problem 3 (10 Marks)
Samples of products are taken and inspected for a process producing certain products. The sample size is constant
of 10. Various numbers of defects are found on 20 samples of inspected products as shown in the following table.
Sample # 1 2 3 4 5 6 7 8 9 10
Number of Defects Found 1 1 1 3 2 1 2 1 1 1
Sample # 11 12 13 14 15 16 17 18 19 20
Number of Defects Found 2 1 2 1 1 1 1 2 1 2
A supplier is to produce large number of parts for one of its customers. The customer requires that the tolerance
for the diameter of the products is 1000 ± 4 and the process capability ratio should be 1.20 or higher. The supplier
measures the diameters of the products using certain measurement equipment while this measurement process has
variations. To determine the variance of the measurement system, 10 manufactured parts are taken and each of the
parts is measured for 3 times. The results of the measurements are shown in the following table. The supplier also
calculated the sample mean and sample standard deviation of the total 30 measurements which are 998.00 and
2.33, respectively. Based on the given data and your calculation, determine:
4
Problem 5 (15 Marks)
1) Identify one or more plans ensuring that, if fraction nonconforming is at 2% or lower, producer’s risk should
be 10% or less. In addition, identify consumer’s risk(s) for such plan(s) if fraction nonconforming is 10%.
2) Identify one or more plans ensuring that, if fraction nonconforming is at 8% or higher, consumer’s risk
should not exceed 10%. In addition, identify the producer’s risk(s) for such plan(s) if fraction
nonconforming is 4%.
3) Briefly explain the procedure of acceptance sampling inspection if the effectiveness of the plan is to be
measured by Average Outgoing Quality Limit (AOQL).
4) Assume that Plan 2 is to be used. Estimate the AOQL associated with this plan.
5) Briefly explain the meaning of AOQL in terms of the quality level of the products after this type inspection.
Pa
ABC Electronics, an electronic manufacturer, purchases large quantity of components from its supplier, XYZ. The
components are supplied in batches with 1,000 units each lot. The Dodge-Romig table (next page) recommends
different sampling plans based on process averages. The figure (next page) presents Type B OC curves for some
of the plans in the table. Based on the given information,
1) Identify the sampling plan for ABC if fraction nonconforming of the components is p=0.5%.
2) Determine the overall quality levels of the components passing the inspection process using the plan
determined in 1) while the actual fractions nonconforming are p=2.0% and p=3.0%, respectively.
3) Assume that the actual fractions nonconforming are p=2.0% and p=3.0%, respectively, calculate the
expected numbers of inspections per lot using the plan determined in 1) above.
4) Identify the LTPD value corresponding to the plan identified in 1) above
5) Briefly explain the meaning of the corresponding LTPD value of the plan in terms of the quality level of
the accepted products.
5
Dodge-Romig Inspection Table – Single-Sampling Plans and Corresponding OC Curves
OC Curves
1
Pa
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0 p
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.2
6
Problem 7 (12 Marks)
MIL-STD-105E standard was followed with single sampling plan to inspect items submitted for inspection in lots
of size 600. The general inspection level is level II and AQL is 0.65%. Following the standard, reduced inspection
was used to inspect Lot 11. Inspection results from Lot 11 to Lot 22 are shown in the following table.
Lot Number 11 12 13 14 15 16 17 18 19 20 21 22
Number of Defective Items 0 0 1 0 2 1 0 2 1 0 0 0
1) Based on the given MIL-STD-105E switching rules and assume that all required conditions are met, identify
the sampling plans (sample size, rejection and acceptance numbers) for inspecting Lots 13, 14, 16, 19 and
21. You may draw the following table in your answer paper and fill up the corresponding information.
Lot 13 14 16 19 21
Inspection Type, Normal (N), Tightened (T) or Reduced (R)
Sample Size
Acceptance number
Rejection number
Lot Accepted (A) or Rejected (Re)
2) Briefly discuss (do not exceed 0.5 page of a standard A4 paper, double space)
2.1) Possible reasons that the inspections are switched between normal, reduced and tightened
2.2) Possible difficulties and ways to overcome such difficulties in implementing this standard
BestMade Manufacturing supplies large quantity of high precision fasteners to major aircraft manufacturers.
BestMade produces about 30 different types or sizes of fasteners with their quantities ranging from 10,000 to
100,000 units each year. Briefly discuss (no more than 1 page of a standard A4 paper, double space) possible
benefits and limitations of using the following statistical tools for quality improvement of BestMade production.
1) Control chart
2) Process capability analysis
3) Acceptance sampling
4) Other relevant QC tools
7
FORMULA SHEET FOR FINAL EXAM
1. Distributions
n
p x (1 − p ) n − x = p x (1 − p )n − x ; µ = np , σ = np(1 − p )
n! 2
Binomial: p( X = x ) =
x!( n − x )! x
e −λ λx
Poisson: p ( x ) = , x = 0,1,2,....
x!
Exponential: f ( x ) = λe − λx , x ≥ 0
Normal: F ( a ) = P( x ≤ a ) = P( z ≤ a − µ ) = ∫
1
z 1 − 2 u2
e dx
σ −∞
2π
2.Point estimators
µ̂ = x , σˆ 2 = S 2 , σˆ = S / c4 , σˆ = R / d 2
3. Average Run Length for control charts
1 1
ARL0 = , ARL1 =
α 1− β
4. Control Charts
p Chart
UCL= p + 3 p(1 − p ) n , Center Line = p , LCL= p − 3 p(1 − p ) n
Type II error for p chart: β = P( pˆ < UCL | p ) − P( pˆ ≤ LCL | p ) = P( D < nUCL | p ) − P( D ≤ nLCL | p )
c Chart
UCL= c + 3 c , Center Line = c , LCL= c − 3 c
Type II error for c chart: β = P ( x < UCL | c ) − P ( x ≤ LCL | c )
u Chart
ui = xi n , u = ∑ xi n × m , UCL= u + 3 u n , Center Line = u , LCL= u − 3 u n
Type II error for u chart β = P ( x < UCL | u ) − P ( x ≤ LCL | u ) = P ( c < nUCL | u ) − P ( c ≤ nLCL | u )
6. Acceptance sampling
Pa p( N − n )
AOQ = or AOQ = Pa p
N
AOQL = max AOQ , ATI = n + (1 − Pa )( N − n )
8
7. Table of critical values
9
9. Military standard MIL-STD-105E
• Switching rules
• Tables
10
11