Nov 1 2011
Nov 1 2011
To be useful, a metaphor must help understanding. For the promoters of Six Sigma to call their certification Black Belt was marketing genius. A more descriptive label might have been Staff Statistician, but what self-respecting manufacturing professional would want to be that? Borrowing a term from Japanese martial arts not only appealed to their fighting spirit, but also gave the impression that an approach developed at Motorola in the US had a connection with Japan, ground-zero of manufacturing excellence. Even in Japan, Black Belt (“Kuroto”) and White Belt (“Shiroto”) have migrated from martial arts to everyday language, to designate respectively a real pro and an amateur.
As a metaphor, Black Belt also made sense because there is a parallel between the Six Sigma and martial arts training programs. Traditional masters in the martial arts of China trained one or two disciples at the Bruce Lee level in a lifetime, just as universities trained only a handful of experts in statistical design of experiments that could be effective in electronics manufacturing. One Karate instructor, on the other hand, can train hundreds of Black Belts, just as a Six Sigma program can teach a focused subset of statistical design of experiments to hundreds of engineers.
Scrum, in software development, is also a sports metaphor, a term borrowed from rugby, which few Americans know. The connection between a rugby scrum and what software people call by the same name, however, is not obvious. A rugby scrum involves the forward players of two teams locked in the pattern of Figure 1.
Figure 1. A rugby scrum
The ball is released in the middle of the scrum and both team try to take possession by kicking it backwards while pushing the other team forwards. It is exciting and bruising to participate in, as well as a great spectacle. For software developers, scrum is an approach to project management illustrated by the status panel in Figure 2.
Figure 2. A software development scrum
It leaves you wondering what plays the role of the opposing team, the ball, or the player positions. In other words, in what way is a rugby scrum a metaphor for this approach at all?
Oct 28 2011
A speaker I once heard on manufacturing metrics started with a quote from football coach Vince Lombardi: “If you’re not keeping score, you’re only practicing.” In a sport, your score or your rank is, by definition, the correct measure of success, and we assume too easily that this kind of thinking crosses over to every human endeavor, from national economies to plant performance or education. In this process, we begin using highly aggregated metrics as if they were physical measurements like mass or speed, and avert our eyes from how these sausages are made.
Following are a few of the egregious examples:
- GDP. Gross Domestic Product (GDP), for example, is in the news every day. If you pollute and spend money to clean up your toxic waste, you contribute more to the GDP than if you produce cleanly. Because of this kind of absurdity, GDP as a metric has been criticized by many economists, including Joseph Stiglitz. In 2009, he even convinced French president Nicolas Sarkozy to seek alternatives. Yet, two years later, the same president is pushing to include in the country’s constitution a “Golden Rule” that caps budget deficits at a percentage of the same flawed GDP!
- IQ. In the US, IQ is still widely treated as a measure of intelligence. On its face, the notion that human intelligence is reducible to a number is an insult to its subject. In fact, all an IQ measures is the ability to take an IQ test. Psychologists recognize this, but many school teachers and the public at large don’t. (See Steven Jay Gould’s The Mismeasure of Man.)
- Food calories. Calories are the most commonly used metric in nutrition. What this number actually represents is the heat generated by drying and burning a food item. But is digestion the same as combustion? Obviously not, for example, for fibers, which cross the human body unchanged. The absurdity of assigning calories to fibers has not escaped one dieter, who questioned it on a Calorie Count forum, and received, among other replies, the following:
Fiber calories are included in nutrition information, but only in come countries. In the US, it is legal to not put in fiber calories because they are not digestible. Therefore, they do not “count” as such. however, if you, like most people, tend to underestimate cals sightly, there is nothing wrong with including them to create a “buffer zone.”
In other words, it makes no sense but you should pretend it does.
Do we behave the same way in the manufacturing world? Yes. For example, many companies measure productivity in terms of Sales/Employee. There is an easy way to boost this metric: outsource all production, close all plants and become a trading company. It is not easy to find metrics for quality, cost, delivery, safety, and morale that are meaningful and cannot be gamed, but it can be done. For overall company productivity, for example, you can use Value added/Employee, where
Value-added = Sales – (Materials + Energy + Outsourced Services)
This is what Peter Drucker called Contributed Value. Value added/Employee is not a perfect metric, but at least it does not provide a perverse incentive to outsource, and the US census bureau publishes statistics on value-added and employment by industry, that are helpful for benchmarking.
Following are a few conditions that a good metric must meet:
- A good metric is immediately understandable. No training or even explanation is required to figure out what it means, and the number directly maps to reality, free of any manipulation. One type of common manipulation is to assume that one particular ratio cannot possibly be over 85%, and redefine 85% for this ratio as “100% performance.” While this makes performance look better, it also makes the number misleading and difficult to interpret.
- People see how they can affect the outcome. With a good metric, it is also easy to understand what kind of actions can affect the value of the measurement. A shop floor metric, for example, should not a be a function of the price of oil in the world market, because there is nothing the operators can do to affect it. Their actions, on the other hand, can affect the number of labor-hours required per unit, or the rework rate.
- A better value for the metric always means better business performance for the company. One of the most difficult characteristics to guarantee is that a better value of a metric always translates to better business performance for the company. Equipment efficiency measures are notorious for failing in this area because maximizing them often leads to overproduction and WIP accumulation.
- The input data of the metric should be easy to collect. Lead time statistics, for example, require entry and exit timestamps by unit of production. The difference between these times then only gives you the lead time is calendar time, not in work time. The get lead times in work time, you then have to match the timestamps against the plant’s work calendar. Lead time information, however, can be inferred from WIP and WIP age data, which can be collected by direct observation of WIP on the shop floor. Metrics of
WIP, therefore, contain essentially the same information but are easier to calculate. (See Little’s Law.)
- All metrics should have the appropriate sensitivity. If daily fluctuations are not what is of interest, then they need to be filtered out. A common method for doing this is to plot 5-day moving averages instead of individual values — that is, the point plotted today is the average of the values observed in the last five days. Daily fluctuations are smoothed away, but weekly trends stand out.
Peter Drucker sold corporate America on the idea that you can’t manage what you can’t measure, and this has led many managers to believe that employees would do whatever it takes to maximize their scores. Given flawed metrics, it if fortunate for the companies that these managers were wrong. If they had been right, all the companies that measure productivity in terms of Sales/Employee would actually have outsourced all production. They didn’t, because metrics are only one of many factors influencing behavior. Most employees, at all level, will not maximize their metrics through actions they feel violate common sense or are inconsistent with their personal ethics.
Oct 27 2011
We have all seen the absurd situation in the featured picture above of a line of customers waiting for taxis while a line of taxis next to them is waiting for customers, with a barrier separating them. This particular instance is from The Hopeful Traveler blog. The cabs are from London, but the same scene could have been shot in many other major world cities.
I am sure we have all encountered similar situations in other circumstances, which may or may not be easy to resolve. One particular case where it should be easy is the restaurant buffet. Figure 1 shows a typical scene in buffet restaurants, with a line of people waiting to get food all on the one side of the table, while food is waiting and accessible on the opposite side.
Figure 1. A typical buffet
I think the fundamental mistake is the assumption that a buffet is like an assembly line, providing sequential access to dishes. This means that you cannot get to the Alo Gobi until the person in front of you is done with the Tandoori. The ideal buffet would instead provide random access, meaning that each customer would have immediate access to all dishes at all times. While it may not be feasible, you can get much closer to it than with the linear buffet. The following picture shows an alternative organization of a buffet in circular islands that is non-sequential.
Figure 2. A buffet island at the Holiday Inn in Visalia, CA
The limitation of this concept is that replenishment by waiters can interfere with customers. To avoid this, you would want dishes to be replenished from inside the circle while customers help themselves on the outside, as in the following sketch:
Figure 3. A buffet island with replenishment from inside
One problem with the circular buffet island, however, is its lack of modularity. You can add or remove whole islands but you cannot expand or shrink an island, which you can if you use straight tables arranged in a U-shape, as in Figure 4.
This buffet island may superficially look like a manufacturing cell, but it is radically different. Its purpose is random access to food as opposed to sequential processing of work pieces, and the materials do not flow around the cell but from the inside out.
Such are the thoughts going through my mind while munching on the Naan at Darbar.
Oct 26 2011
Data mining, in general, is the retrieval of information from data collected for a different purpose, such as using sales transaction histories to infer what products tend to be bought together. By contrast, design of experiments involves the collection of observations for the purpose of confirming or refuting hypotheses.
This perspective on data mining is consistent with the literature in expressing purpose, but most authors go further. They include in their definitions that data mining is done with computers, using large databases and specific analytical tools, which I think is too restrictive. The tools they list are the ones they have found useful in analyzing the behavior of millions of users of search engines or commerce websites, and they are not obviously applicable in other areas, such as manufacturing.
During World War II, British analysts used the serial numbers of captured or destroyed German tanks to estimate the numbers produced. Because serial numbers were not attached for this purpose, it was data mining. It used clever statistical models but, obviously, no computers.
Today, PhD-level data miners at Google, eBay, or Amazon sift through the page views and click-throughs of millions of users for clues to patterns they can use. The data, automatically collected, is accurate and collected by the terabytes every day. This “big data” requires parallel processing on clusters of computers and lends itself to the most advanced analytical tools ever developed.
Compared to this fire hose of data, what manufacturing produces is a trickle. In a factory, the master data/technical specs, plans and schedules, status of operations and work in process, and the history of production over, say, 12 months, usually adds up to a few gigabytes. It doesn’t fit on one spreadsheet, but it often does on a memory stick. On the other hand, much of it is still manually generated and therefore contains errors, and it is often structured in ways that make it difficult to work with.
Even if manufacturing companies could hire the data miners away from their current jobs, their experience with e-commerce or web search would not have prepared them well for the different challenges of manufacturing data mining.
There is an opportunity for data mining to contribute to competitiveness in manufacturing, but the approach must start from the needs. It must not be an e-commerce cure in search of manufacturing diseases.
Oct 25 2011
The first draft of my book Working with Machines contained a chapter that was a post-mortem on two obsolete machines, which was cut on the grounds that, unlike all other chapters, it was not actionable for the reader.
Its abstract is as follows:
The steam locomotive and the typewriter are icons of the industrial age, and their parallel histories show different aspects of the human experience of working with machines. The steam locomotive is fondly remembered; the typewriter, all but forgotten except for the QWERTY keyboard. The steam engine participated in the development of every industrial economy, but the typewriter played no major role in Japan. The typewriter did not demonstrably improve the productivity or quality of office output, but was adopted only because of its image of modernity.
Locomotive driver was a prestigious position for a manual laborer, but typist never was. Compared to electrics and diesels, the steam locomotive had a cab that was exposed to the elements and to the heat of the firebox and therefore uncomfortable, difficult to operate, and dangerous. Yet engineers and firemen preferred it to the tedium and loneliness of modern locomotives. Automatic machines that require human attention only when they malfunction are also in airplanes and in manufacturing plants, challenging the job designer to keep the operator alert and used efficiently.
As the typewriter prints one keystroke at a time, typists were always busy with a single machine and determined both its productivity and output quality. Typists worked in comfortable places, but under pressure, and faced the long-term hazards of sedentary work. The typewriter’s main legacy is that a society can make a long-term investment in machines whose tangible benefits do not obviously exceed their costs.
Click here for a pdf file of the entire chapter.