Monday, March 28, 2022

Statistics and Industrial Engineering


Lesson 401   of  Industrial Engineering ONLINE Course -  IE Statistic and IE Six Sigma Module.

Statistics and Industrial Engineering
Author: Narayana Rao

F.W. Taylor himself advocated maintaining of records and data for decision making. The other industrial engineering pioneers also promoted record keeping and data analysis. As sampling based  decision making became more robust, industrial engineers promoted it as a productivity improvement initiative and imperative. One of the prominent areas of application is statistical quality control. Sampling was also used in work measurement and work sampling technique was developed in industrial engineering. Now six sigma, a statistics based technique is being promoted by the IE profession.




F.W. Taylor has indicated that data collected for machine shop will be in thousands of pages. Harrington Emerson included records in his book 12 Principles of Efficiency. Their contemporary, professor of industrial engineering, Diemer wrote:

Department of Records.
"It is primarily a research and advisory department the results of  whose investigations and whose recommendations are brought up  at such meetings of department heads and others as may have been  predetermined. It is the duty of the record department to see that  records kept by various departments are not merely kept and stored  away, but that from each set of records is secured a method of most effective analysis so that the records of the past may be compared  with records of the present and conclusions may be drawn as to future  action. The individuals engaged in this department must be experts  in theory of accounts, the science of statistics, the art of graphical  presentation and cost accounting. The tendencies and facts indicated by an analysis of the records must be brought forcibly  to the attention of all individuals whose actions based on experience  and intuition differ from the action indicated by an analysis of figures,  records and statistics."

Reference: Factory Organization in Relation to Industrial Education
Author(s): Hugo Diemer
Source: The Annals of the American Academy of Political and Social Science, Vol. 44, The
Outlook for Industrial Peace (Nov., 1912), pp. 130-140

Industrial engineering has taken up the responsibility of using statistics to make processes in organizations efficient. May be Walter Shewart is the first statistician to develop a systematic method for applying the concepts and methods of statistics to industrial process control problems and industrial engineering has adopted statistical process control as a method to be installed in companies through IE department.

The role of statistics as a tool of management
J. M. Juran
Statistica Merlandica
Volume 4, Issue 1‐2, February 1950, Pages 69-79
First published: February 1950 
† *Paper presented at the 26th session of the International Statistical Institute in Bern, September 1949. Reprinted with permission of the author and of the ISI.

Growth of the mass production industries has posed new and complex problems in industrial management. Scientific solution of these problems necessitates statistical analysis of the vast quantities of data generated in these industries as a by‐product.

Improvements bordering on the spectacular have been achieved in selected instances of industrial applications of statistical analysis. Quality control and market research afford two such instances.

The professional statistical societies can do much to aid the greater utilization of statistics in industry by:

(a). organizing in each society a major division to deal with the problems of statistics in industry.

(b). sponsoring joint meetings with societies of managers, industrial engineers, and others interested in industrial statistics.

Important applications of statistics in industrial engineering: Work Sampling, Statistical Quality Control, Design of Experiments to improve productivity, Six Sigma
_______________________________________________________________________

Variability

No  two objects in the world around us, nor any two actions performed by the same or by different individuals, are exactly identical. Precision machine parts produced in quantity by the same operator busing identical tools and equipment will, upon examination show a definite variability.

Manufacturers try to reduce the variability of their output. The complete elimination of  variability in production is usually not feasible, and would be entirely uneconomical even if feasible. Instead, the manufacturer's philosophy is based on a tolerable, statistically predictable, level of imperfect product.

Source:   Siegmund Halpern, The Assurance Sciences, Prentice-Hall, Inc,. Englewood Cliffs, New Jersey, 1978,p.66.

Quality control enables us to ascertain sudden or gradual changes in product variability (or establish trends) to permit the institution of timely corrective action that will avoid production of costly scrap.

________________________





Remembering Walter Shewhart (Quality Magazine, March 2, 2009)
http://www.qualitymag.com/Articles/Web_Exclusive/BNP_GUID_9-5-2006_A_10000000000000540505

THE PHILOSOPHY OF SHEWHART'S THEORY OF PREDICTION
Dr Mark Wilcox, Centre for Business Performance; Cranfield School of Management, Cranfield University, Cranfield, United Kingdom. MK430AL
http://www.flowmap.com/documents/shewhart.pdf

Multivariate Quality Control - Historical Perspective
http://www.opf.slu.cz/vvr/akce/turecko/pdf/Firat.pdf

Shewhart’s Charts and the Probability Approach
Henry R. Neave and Donald J. Wheeler
© 1996
http://www.spcpress.com/pdf/Wheeler_Neave.pdf


Variation through Ages,  Quality Progress, Dec 1990 (interesting article)
http://www.apiweb.org/VariationThroughAges.pdf



Book: Industrial Statistics: Practical Methods and Guidance for Improved Performance

Anand M. Joglekar

ISBN: 978-0-470-49716-6 April 2010 288 Pages

TABLE OF CONTENTS
PREFACE.
1. BASIC STATISTICS: HOW TO REDUCE FINANCIAL RISK?

1.1. Capital Market Returns.

1.2. Sample Statistics.

1.3. Population Parameters.

1.4. Confidence Intervals and Sample Sizes.

1.5. Correlation.

1.6. Portfolio Optimization.

1.7. Questions to Ask.

2. WHY NOT TO DO THE USUAL t-TEST AND WHAT TO REPLACE IT WITH?

2.1. What is a t-Test and what is Wrong with It?

2.2. Confidence Interval is Better Than a t-Test.

2.3. How Much Data to Collect?

2.4. Reducing Sample Size.

2.5. Paired Comparison.

2.6. Comparing Two Standard Deviations.

2.7. Recommended Design and Analysis Procedure.

2.8. Questions to Ask.

3. DESIGN OF EXPERIMENTS: IS IT NOT GOING TO COST TOO MUCH AND TAKE TOO LONG?

3.1. Why Design Experiments?

3.2. Factorial Designs.

3.3. Success Factors.

3.4. Fractional Factorial Designs.

3.5. Plackett–Burman Designs.

3.6. Applications.

3.7. Optimization Designs.

3.8. Questions to Ask.

4. WHAT IS THE KEY TO DESIGNING ROBUST PRODUCTS AND PROCESSES?

4.1. The Key to Robustness.

4.2. Robust Design Method.

4.3. Signal-to-Noise Ratios.

4.4. Achieving Additivity.

4.5. Alternate Analysis Procedure.

4.6. Implications for R&D.

4.7. Questions to Ask.

5. SETTING SPECIFICATIONS: ARBITRARY OR IS THERE A METHOD TO IT?

5.1. Understanding Specifications.

5.2. Empirical Approach.

5.3. Functional Approach.

5.4. Minimum Life Cycle Cost Approach.

5.5. Questions to Ask.

6. HOW TO DESIGN PRACTICAL ACCEPTANCE SAMPLING PLANS AND PROCESS VALIDATION STUDIES?

6.1. Single-Sample Attribute Plans.

6.2. Selecting AQL and RQL.

6.3. Other Acceptance Sampling Plans.

6.4. Designing Validation Studies.

6.5. Questions to Ask.

7. MANAGING AND IMPROVING PROCESSES: HOW TO USE AN AT-A-GLANCE-DISPLAY?

7.1. Statistical Logic of Control Limits.

7.2. Selecting Subgroup Size.

7.3. Selecting Sampling Interval.

7.4. Out-of-Control Rules.

7.5. Process Capability and Performance Indices.

7.6. At-A-Glance-Display.

7.7. Questions to Ask.

8. HOW TO FIND CAUSES OF VARIATION BY JUST LOOKING SYSTEMATICALLY?

8.1. Manufacturing Application.

8.2. Variance Components Analysis.

8.3. Planning for Quality Improvement.

8.4. Structured Studies.

8.5. Questions to Ask.

9. IS MY MEASUREMENT SYSTEM ACCEPTABLE AND HOW TO DESIGN, VALIDATE, AND IMPROVE IT?

9.1. Acceptance Criteria.

9.2. Designing Cost-Effective Sampling Schemes.

9.3. Designing a Robust Measurement System.

9.4. Measurement System Validation.

9.5. Repeatability and Reproducibility (R&R) Study.

9.6. Questions to Ask.

10. HOW TO USE THEORY EFFECTIVELY?

10.1. Empirical Models.

10.2. Mechanistic Models.

10.3. Mechanistic Model for Coat Weight CV.

10.4. Questions to Ask.

11. QUESTIONS AND ANSWERS.

11.1. Questions.

11.2. Answers.

APPENDIX: TABLES.

REFERENCES.

INDEX.

https://www.wiley.com/en-us/Industrial+Statistics%3A+Practical+Methods+and+Guidance+for+Improved+Performance-p-9780470497166

INDUSTRIAL STATISTICS

Course Developers
D. K. Jain & R. Malhotra
http://ecoursesonline.iasri.res.in/course/view.php?id=484

Industrial Statistics - Guidelines and Methodology - Unido
https://www.unido.org/resources/publications/cross-cutting-services/industrial-statistics-guidelines-and-methodology

Last edited on Knol: 22 Jul 2010
Exported : 26 Nov 2011 to this blog

Original URL: http://knol.google.com/k/-/-/  2utb2lsm2k7a/ 2768


Key Words: Statistics, Industrial Engineering and Efficiency
__________________________________________________________________________


Updated 28.3.2022,   9 August 2021,  4 August 2019,  14 July 2016,  23 July 2012

No comments:

Post a Comment