Comparing Handbooks: Maynard, Salvendy, Badiru, NITech

Industrial Engineers most often cite Maynard’s and Salvendy’s handbooks, both last updated in 2001. The most recent English-language handbook I know of is Badiru’s, whose 2nd edition came out in 2013. NITech is the Nagoya Institute of Technology (名古屋工業大学). Since 2007 NITech has been running a 6-month Plant Manager Training School (工場長養成塾) program once a year, including lectures, plant visits, and projects. This program has a companion handbook last updated in 2015. It’s focused on plant managers rather than IE’s but I included it here because it represents a different approach. The most recent publication I checked out is the 2019 Industrial Engineering Body of Knowledge (IEBoK) from the IISE but it is only an outline, with a bibliography on each topic.

In principle, a comparative analysis of the four handbooks requires reading 6,500 pages of English text and 244 pages in Japanese, with massively overlapping content. Robert Caro’s editor in the 1960s instructed him to “Turn every page,” and he did, in Lyndon Johnson’s voluminous archives. He wrote a definitive biography as a result. Much as I respect Caro’s dedication, I cannot practically follow his lead. While I normally don’t review books I haven’t read, these are reference books, like dictionaries or encyclopedias. You can review them without first reading them cover-to-cover. Amazon buyers post reviews of Webster’s after they “looked up three obscure words to test it out.”

In this spirit, what we can do with this much material is start from the most readily available information: who the authors were, the topics they chose to include, and the volume of information on each, as measured, imperfectly, by page count. It tells us what is on the menu in each book. Then we spot-check a sample of topics, focussing on the differences in treatment among the handbooks.

Top-Level Comparison

The most obvious characteristics of handbooks are as follows:

TitleIndustrial Engineering Handbook 5th EditionHandbook of Industrial Engineering, 3rd EditionHandbook of Industrial and Systems Engineering, 2nd EditionPlant Manager Training School Handbook (工場長養成塾ハンドブック)
Cover
LanguageEnglishEnglishEnglishJapanese
Vintage
200120012013
2015
Contributors175185678
Page count2,6882,8361,476244
Pages/contributor15152331
Sellers rank 546,8852,178,0832,168,923272,316 (Amazon Japan)
Price $83$369$186¥1,540 Price/page$0.03$0.13$0.12¥6.3

For the three Industrial Engineering handbooks, the sellers rank and price data from Amazon for print editions as of 12/29/2019. For the Japanese handbook, the data are from the Japanese Amazon site.

Note that Maynard’s handbook is 4 times cheaper per page than Salvendy’s and Badiru’s. This may explain in part why its Amazon rank is higher. All three handbooks are available in electronic form. As mentioned in my previous post, this is the only form in which they qualify as “conveniently carried.” Badiru’s handbook has barely 1/3 as many authors as the other two, for 55% and 57% of the page count. Each author contributed on the average 50% more in Badiru’s than in Maynard’s or Salvendy’s handbooks. This being said, Badiru’s handbook is still a massive, collective effort. Unlike the American handbooks, NITech’s print version is small enough to fit in a jacket pocket.

The three American handbooks all include a discussion of the Industrial Engineering profession, which was the subject of a previous post. One out of five chapters in the Japanese handbook is about the role of the plant manager and the knowledge required to fill it.

Author Affiliations and Locations

For all four handbooks, the authors are mostly academics and consultants. Industrial Engineering consultants dominate with Maynard; academics, with Salvendy and Badiru. The authors of the NITech handbook are all Japanese but do not list their affiliations. The following tables summarize the handbook contributors by professional category and location. The Country columns show where they work, not necessarily where they are from.

Contributors to Maynard’s Industrial Enginering Handbook

The editor, Kjell Zandin,  who died in 2011, was Swedish and worked in Sweden and Germany for the H.B. Maynard and Co. consulting firm prior to moving to the US in 1975. Besides the Handbook, Zandin is best known for MOST, a refinement of the MTM predetermined time system developed by Maynard in the 1940s. MOST works directly on sequences of the elementary motions in MTM. The automotive industry uses both systems.

The Maynard handbook is co-sponsored by JMA, the largest manufacturing consulting firm in Japan, founded in 1942, whose past members include Shigeo Shingo and TPM creator Seiichi Nakajima. All but two of the 14 Japanese co-authors are affiliated with JMA.

The summary in the following table shows that Maynard’s handbook was primarily written by consultants working in the US. That Japan should come in second is no surprise but the strong showings from Sweden and Canada are.

CategoryCount%Cum %CountryCount%Cum %
Consulting8045%45%USA12470%70%
Manufacturing1710%82%Sweden116%85%
Engineering53%88%Germany42%92%
Software53%91%Israel32%94%
Energy21%92%UK32%95%
Healthcare21%93%Czech Rep.21%97%
Logistics21%94%Mexico21%98%
Construction11%95%Austria11%98%
Legal services11%96%Finland11%99%
Real Estate11%96%Slovakia11%99%
Retail11%97%Switzerland11%100%
Services11%97%
Other53%100%

Contributors to Salvendy’s Handbook of Industrial Engineering

Until the last edition of this handbook came out in 2001, Gavriel Salvendy was a professor of Industrial Engineering at Purdue University. He went on to head the IE department at Tsinghua University in Beijing, China and is now a Distinguished Professor at Florida Central University. Besides this handbook, Salvendy worked in human factors and ergonomics and edited another handbook on this subject. Educated in Israel and the UK in the 1960s, he has since had a broad international career, and has been mostly based in the US.

Salvendy’s handbook was primarily written by academics working in the US. Among international contributors, Germany eclipsed Japan as the 2nd largest group. Chinese authors, absent in Maynard’s handbook, made contributions from the Mainland (“P.R. China”), Taiwan (“R.O. China”), and Hong-Kong.

CategoryCount%Cum %CountryCount%Cum %
Consulting158%83%Germany2815%85%
Logistics53%85%Israel42%87%
Government42%88%Taiwan42%89%
Software32%91%Japan32%93%
Engineering21%92%China32%94%
Professional Society21%94%Greece21%95%
Bank11%94%Hungary21%96%
Healthcare11%95%The Netherlands21%97%
Non-Profit Organization11%95%Chile11%98%
Retail11%96%Hong Kong11%98%
Other84%100%Philippines11%99%
Singapore11%99%
Turkey11%100%

Contributors to Badiru’s Handbook of Industrial and Systems Engineering

The editor of the newest handbook, Adedeji Badiru, is Dean of Graduate School of Engineering & Management at the Air Force Institute of Technology (AFIT). His education and career have been in American universities. Abroad, he is a Fellow of the Nigerian Academy of Engineering and has received several awards from Nigeria. In addition to this handbook, he has written or edited many others, on military industrial engineering and on project management. American academics dominate Badiru’s group of co-authors even more than Salvendy’s.

CategoryCount%Cum %CountryCount%Cum %
Consulting23%94%India34%84%
Energy12%95%Nigeria34%88%
Goverment12%97%Taiwan34%93%
Manufacturing12%98%UK23%96%
Singapore11%99%

Contributors to NITech’s Plant Manager Training School Handbook

This handbook has the institution of the NITech Plant Manager Training School as “editor.” Only the postface is signed by 8 individuals, who do not include any affiliation information.

Topics covered

Maynard’s Handbook Contents

The following table shows the parts of the Maynard handbook, sorted by decreasing page count, followed by comments on the elements that stand out:

Subject # of Pages% of totalCum %
Ergonomics and safety24810%10%
Productivity, performance, and ethics2229%18%
Tools, techniques, and systems2179%27%
Work measurement and time standards2028%35%
Industrial engineering:past, present, and future1847%42%
Forecasting, planning, and scheduling1767%49%
Manufacturing technologies1586%55%
Logistics and distribution1446%61%
Facilities planning1405%66%
Statistics and operations research and optimization1345%72%
Engineering economics1305%77%
Product design and quality management1265%82%
Compensation management and labor relations1064%86%
Government and service industry applications1024%90%
Work analysis and design964%94%
Maintenance management883%97%
Information and communication management743%100%

Ergonomics and Safety

Maynard’s handbook devotes 248 pages to Ergonomics and Safety — more than to any other subject — but the section is mostly about ergonomics. In manufacturing, it is a topic that one should not treat as part of stand-alone workstation design but within the context of production line design and operation by teams.

For example, you may improve the ergonomics of an individual workstation by an adjustable platform for operators of different heights. On the other hand, a production line may require an operator to move between stations, which makes adjustable platforms at each station impractical. And since the flow of materials between stations work surfaces them to be at the same height, these heights cannot be adjusted.

Job rotations are also a consideration. If operators switch positions every two hours, then you must analyze repetitive stress issues over two hours, not an entire shift.

Overall the line designers must accommodate the population of operators, with the mode of operation. There is more to it than designing for an individual at a single station.

Ethics

Maynard’s handbook is the only one to discuss Ethics, in Chapter 2.7, by Larry J. Shuman and Harvey Wolfe. An IE faced with an ethical dilemma, however, would find no guidance there on how to resolve it. The authors “wish to sensitize the engineer or engineering student.” They show examples of problems, reproduce the ABET “Canon of Ethics,” quote Cicero, and outline probabilistic risk assessment but provide no guidelines or recommendations on how to be an ethical IE.

Salvendy’s Handbook Contents

The same table for Salvendy’s handbook has Manufacturing and Production Systems topping the chart with 312 pages, as part of a section entitled Technology, which groups it with Information Technology and Service Systems.

Subject # of Pages% of TotalCum %
Manufacturing and Production Systems31212%12%
Information Technology2469%21%
Human Factors and Ergonomics2269%30%
Service Systems2168%38%
Quality2148%46%
Organization and Work Design1727%52%
Planning and Control1666%59%
Systems and Facilities Design1626%65%
Probabilistic Models and Statistics1526%71%
Supply Chain Management and Logistics1365%76%
Optimization1305%81%
Computer Simulation1144%85%
Manpower Resource Planning1124%89%
Economic Evaluation1124%93%
Product Planning703%96%
Industrial Engineering Function and Skills602%98%
Project Management422%100%

Manufacturing and Production Systems

Under this title, I expected to find a detailed account of the Toyota Production System (TPS), explaining how line designs dovetail with supply chain management, quality, and human resource management.

I also expected to read about the production systems of other carmakers and of companies making electronics, airplanes, furniture, frozen foods, paints, etc.  A contrast with earlier approaches like Ford’s Mass Production, GM’s Flexible Mass Production or Junker’s Taktsystem would have provided perspective.

Instead, the 12 chapters cover a range of topics in which I see no obvious throughline:

• Enterprise modeling, at a level that is only of interest to software engineers.
• ERP, with no use of the tools introduced in enterprise modeling.
• Near-net shaped processes, a metalworking technology of interest if you make gears but not if you make frozen lasagna.
• Environmental engineering, from the perspective of compliance with US regulations.

The only mention of TPS is in Chapter 17, about Just-in-Time, Lean Production, and Complementary Paradigms.

The reason the modeling tools of Chapter 9 are not used in Chapter 11 is that different authors wrote them. It does, however, make the handbook read like a mystery story with a gun on the wall that doesn’t go off. Chapter 9 foreshadows the use of the tools in later chapters. Then it doesn’t happen. By contrast, Manufacturing Systems Analysis, with Application to Production Scheduling introduces analytical tools in the first half of the book and then systematically uses them in the second half to compare different approaches to production scheduling. The work of a single author usually has this kind of consistency.

Information Technology

Industrial Engineering concepts retain their relevance for decades, when not centuries. On the other hand, anything you write today about information technology (IT) is likely to be obsolete in five years. Accordingly, this section presents a vision of where information technology was headed at the turn-of-the-century. In addition, it appears to target engineers who design or select and implement enterprise systems.

It’s not what most industrial engineers do with IT.  Usually, the company’s information systems are a given and IEs wrangle useful information out of them. For example, as part of a project, an IE may want to mine receiving transactions for trustworthy suppliers or extract the information embedded in “smart” part numbers. This section does not address such challenges.

Service Systems

As for production systems, the chapters cover a variety of topics but do not synthesize what a service system is. This would list the components and describe how they interact with each other, with customers, suppliers, complementers, or competitors. I would have expected descriptions of one or more successful service organizations and discussions of how their concepts can be generalized.

This part ends with five chapters on the application of IE to health care, financial asset management, retail, transportation, and hospitality (“hotels and restaurants”). The chapter on health care is focused on the critically dysfunctional US system when other nations achieve better outcomes at lower costs, as shown in the following table:

CountryLife expectancy at birth in Years (2018)Per capital health care expenses/year (2015)
US78.9$9,536 Japan84.5$3,733
Italy83.4$2,700 Spain83.4$2,354
France82.5\$4,026

The chapter on financial asset management reads more like an introduction to the math of investments than a description of a system rendering services to individuals and organizations. The chapters on retail, transportation, and hospitality come closer to actually describing systems.

This section also includes two chapters that aren’t about service industries:

• Mass Customization (Chapter 25), is primarily a manufacturing concept.
• Client/Server Technology (Chapter 26) is a software architecture concept that has nothing to do with the business of delivering services. It belongs in IT.

Operations Research (OR) rises to the top spot in Badiru’s handbook. This is up from 10th position in Maynard’s and 11th in Salvendy’s. In Maynard’s OR shares 134 pages with statistics, with 130 pages, under the title of Optimisation. Both Maynard’s and Salvendy’s handbooks devote 5%  of their pages to OR. In Badiru’s shorter book, the 236 pages on OR account for 19% of the total.

Subject # of Pages% of TotalCum %
Operations research, queuing, logistics, and scheduling23619%19%
Management, information engineering, and decision making21417%36%
Fundamentals of industrial engineering16813%49%
Human factors and ergonomics15412%61%
Manufacturing, production systems, and ergonomics14011%72%
Economic and financial analysis13411%83%
Fundamentals of systems engineering1068%91%
Safety, reliability, and quality968%99%
General introduction141%100%

Operations Research

The part’s complete title lumps OR with queuing, logistics, and scheduling. The theories of queuing and scheduling, however, are part of OR, as the information dimension of logistics. The title implies that the part treats the whole of OR, including these subtopics. In fact, it includes nothing but these subtopics. It is not an overview of OR as commonly taught. For example, it doesn’t introduce mainstays of OR like linear, dynamic and integer programming.

That the OR section is the tail end of the book, pp. 1029 to 1264, suggests the authors view it as an advanced topic. That they devote more space to it than to any other subject tells us they view it as central. They view work measurement, work sampling, human factors,…, as foundations. OR is what they build on top of these foundations.

As discussed below, there is a disconnect between academia and the manufacturing world about OR. Today, it is the dominant research interest of IE professors nationwide but, essentially, it never comes up in factories.

NITech handbook contents

It has five chapters of about 50 pages each:

1. How should a plan manager be?
2. Making a factory appealing.
3. Achieving today’s production.
4. Improving for tomorrow.
5. Developing people for the future.

The pages are small and of unusual size: 7 in x 4 in. They organized in 2-page spreads about a single topic, as in the following example:

These pages are excerpted from Chapter III, about Achieving Today’s Production. Within this chapter, it is in a section about A Day in the Life of a Plant Manager, subsection about Preparing to Launch Production. The full translation of this spread is as follows:

This handbook contains no philosophy or theoretical background. It is short, prosaic, and relentlessly practical. It is surprising in the product of an academic institution, albeit a stone’s throw from Toyota’s headquarters.

As a conveniently carried reference, it has a legitimate claim to be called a “handbook.” In the US, however, it is closer in spirit to the “for Dummies” series, or books called “Quick Start Guides” or “Cheat Sheets.” The closest existing title is Operations Management for Dummies but it’s not about managing factories.

Conclusions

The Industrial Engineering handbooks fail to clarify Industrial Engineering as a discipline. They don’t explain either how it fits in with Systems Engineering and Operations Research. The IISE, Body of Knowledge lists the duality theory of linear programming among what an IE must know. I don’t recall the topic ever coming up in conversation with an actual IE.

OR in Manufacturing?

Given the growing prominence of OR in academic Industrial Engineering reflected in the American handbooks, we need to take stock of what contributions OR has actually made to this field. Manufacturers do use some simple results from OR, usually without attribution to OR:

• Little’s Law relates the steady-state means of throughput, inventory, and lead times in operations and is used, among other things, to size Kanban loops.
• Kingsman’s Formula, with very few restrictions, says that, when a single-server queue saturates, its waiting time grows like the inverse of the server’s availability. If its 98% busy, its waiting time will be twice as long as if it were 96% busy.
• Bucket Brigades are a method for organizing the assembly of custom-configured products in a self-balancing manner invented by John Bartholdi at Georgia Tech. It is used in the Toyota Sewing System for car upholstery, at Timbuk2 Designs in luggage assembly, to make sandwiches at Subway restaurants, and for picking in many warehouses.
• The Prisoner’s Dilemma is a classic example from game theory that I used in Lean Logistics as a framework to explain why collaborative customer-supplier relationships are difficult to sustain.
• Simulations. Manufacturers use simulations to anticipate the behavior of various systems prior to implementing them.

OR plays almost no role where you would most expect it to, Production Planning and Scheduling.

Optimization Versus Improvement

In spite of its modest contributions to Manufacturing, OR has influenced engineering and management thinking by promoting optimization, over improvement. One of the earliest posts in this blog was about how a factory can always be improved. With an optimization mindset, you seek the one best way. By definition, there is no way to improve on it. The “one best way,” however, only exists in a mathematical model, not in an actual shop. There is nothing wrong with a mathematical model. Just remember that you can always refine or change it to open new opportunities.

With an improvement mindset, every improvement sets the stage for the next one. Pointing out improvement opportunities in an assembly line, I have heard managers respond “We’ve already optimized this.” It meant that they had no intention of doing anything about this line.

Obstacles to OR in Manufacturing

Historically, the spread of OR techniques in manufacturing has faced several limitations:

1. Underlying assumptions.
2. Detailed data requirements.
3. Computing power requirements.

These issues are illustrated below in the examples of safety stock calculations, milk run planning, and setup sequencing,

Safety Stocks

Inventory management is a major topic in OR and OR has produced a formula to set safety stock levels that is the subject of Safety Stocks: Beware of Formulas, the 2nd most popular post in the history of this blog. It contains an example of an actual misapplication of the formula, based on unrealistic assumptions about consumption and replenishment lead times.

If these assumptions had been valid, the formula would have needed not only means but also standard deviations for consumptions and lead times for each item, and estimating them requires a level of access to raw transaction data that was not provided by factory IT. The engineer plugged in an arbitrary 20% of the mean for the unavailable standard deviations.

Once you have the data, the formula doesn’t require more than an electronic spreadsheet. In the example, however, the engineer put exponents on the wrong variables in the Excel formula. Assuming the formula produced correct thresholds for a few thousand items, these thresholds still needed to be uploaded to the ERP system in order to trigger replenishment orders, and this would have to happen every time the organization reviews and updates its master data.

Milk Run Planning

To use milk runs in collecting parts from suppliers, you need to assign groups of suppliers to trucks and map routes for each truck, and it sounds like a perfect challenge for OR. It is currently an active research area in OR, with most publications in the 2010s. The concept of a supplier milk run goes back, at least, to the German aircraft industry in World War II, so why did it take so long for OR to work up an interest in this topic?

For decades, practitioners have muddled through the job of partitioning a set of 200 geographically distributed suppliers into clusters of 5 or 6 to  visit on a milk run. Perhaps genetic algorithms can partition the suppliers faster and better but it’s too early to tell. Once you have selected a group of suppliers, routing a truck through all the suppliers in the group is the OR classic Traveling Salesman Problem (TSP). Research on the TSP seems focused on solving ever-larger problems. On one supplier milk run, you visit 6 suppliers, not 30,000, and you can evaluate the possible routes by brute-force enumeration of the 720 possible sequences. It works if you want to minimize total truck mileage because the distance between suppliers does not change. It doesn’t if you want to minimize the total time, because traffic and weather conditions affect it randomly.

Online mapping services provide travel time estimates between two stops based on current traffic conditions. Multi-stop route planning services then use this data in real-time to generate routes that minimize time or fuel consumption. It exists and it’s not particularly difficult to understand or use but it’s not the direction the OR literature goes into.

Setup Sequencing

Formally, the problem of sequencing setups in a machine to cycle through multiple operations is the Traveling Salesman Problem, with the machine as the “Salesman,” the operations as “cities,” and the changeover times as “travel times.” In principle, the changeover times are deterministic but, unless the machine has undergone a setup time reduction project (SMED), they are actually random. Not only do they involve variable times to fetch tooling that should have been prepared ahead of time but starting up the new operation requires multiple attempts in varying numbers.

In addition to making changeover times short and fixed, SMED projects are often accompanied with a re-allocation of operations to machines, resulting in a specialization of the machines and a drastic reduction in the number of different operations done with each. If you are down from 50 operations to 8 on a machine, you are again in the domain where brute-force enumeration of all the 40,320 possible sequences is feasible.