Problems and Tactics Table: Human Fit
Since products of industry are to serve people, the fit between people and products is an important issue. The desire for such fit is certainly not a new one. In fact, a tremendous amount of time and effort has been applied to this area. The field of human factors engineering, complete with the Human Factors and Ergonomics Society, focuses upon the relationship between people and machines. The term ergonomics applies to this area, along with biotechnology, which is still defined in some dictionaries as the study of human-machine compatibility, although it has been co-opted by the field of genetic engineering. More recently, the term human interface has been used in describing this relationship. Divisions of technical societies such as the Institute of Electrical and Electronic Engineers (IEEE) focus on this area, while journals including the Journal of Human Engineering, the Journal of Engineering Psychology, and the International Journal of Man-Machine Studies report on new understanding of this relationship. You can easily find large numbers of books that contain concepts and data of value in making products that properly fit people.
Still, we have problems with human fit.
Many people have trouble removing the lids from jars. The majority of airplane crashes are attributed to pilot error. Women’s high-heeled shoes permanently deform their toes. We worry about carpal tunnel syndrome. Comedians amuse us parodying people who cannot program their DVRs. We can’t read the labels on our electronic equipment controls. What is going on?
Good products must fit people, and there is tremendous room for improvement. Take a few minutes and think of products that do not fit you—that do not match your body, sensory system, or mind. You should have no trouble doing this, for the world is full of products that ask us to greatly inconvenience ourselves in order to interact with them. Designers should know better than to do many of the things they do when designing a product. Even worse, they often do know better. I taught a course in human engineering to design majors in the engineering school at Stanford. The class was frustrating to teach because the material we covered seemed obvious to the students. Of course an auditory warning signal was better than a visual one if the operator’s task required visual attention. Obviously, chairs should be designed to be kind to the skeletal system. Who would design a control panel without grouping controls for related functions and keying controls to related displays? To my amazement, the students, after reprimanding me for the commonsense nature of the course, would then go straight to their design jobs in industry and proceed to violate the very principles they considered so obvious.
Designers often become so distracted by considerations other than human fit that it does not receive the attention it deserves. Product function, cost, appearance, and reliability may dominate the designer’s attention. Tight schedules or budgets do not encourage prototypes and usage tests. Even worse, designers may design for themselves: the petite, young, physically fit, right-handed designer may not think about large, older, left-handed users with physical disabilities. In addition, designers are used to taking advantage of the tremendous adaptability and flexibility of humans. People are capable of accommodating to ridiculously poorly designed products and do so surprisingly quickly with a minimum of complaints. Trouble can occur, however, if users have to accommodate to these products too often, rapidly, accurately, or too long. Furthermore, why should we have to?
My classes were often televised, and individual microphones were provided for students in the classroom to provide clearer audio to the remote students. In the room I often used, the students were supposed to grab a small microphone from a rack on the back of the seat in front of them, bring it to their mouth, and press a button before they spoke. The microphones hung so that the side into which the students were to speak faced them. This setup looked all right until you realized that when you grab something at arm’s length and bring it to your mouth it rotates 180 degrees. The normal motion therefore brought the back of the microphone to the mouth of the student. The microphones also stuck in their holders. The students could learn to use the microphones, but why should they have to? Needless to say, most of them found it easier either to ignore the microphones when they spoke or not to engage in discussion.
There are also reasons of tradition and culture that cause people to ignore improving the fit of products. A simple example is the Western bathroom. Many years ago, the architecture department at Cornell University did a study on this topic and produced a wonderful book, The Bathroom, criticizing almost every aspect of the bathroom. Tubs are unsafe, difficult to enter and leave, and not efficient in cleaning. Toilets are anatomically incorrect. Lavatories do not allow one to wash one’s hair without washing the remainder of the room. Mirrors fog up. The study is supported by a great amount of data, including a statistical analysis of male urine streams, concluding that an appreciable amount of them miss the toilet. There also are pictures of people groping for their shampoo while trying to control a rubber hose with a sprinkler on the end. It is, all in all, a damning indictment.1
People don’t want to think about bathrooms. Nor do they want to experiment. We cling to the bathroom in which we were reared. One of the traumatic memories in the minds of many “Western” people is their first experience with a squat toilet. A bidet is a useful device, but it has never met much of a market in the United States because we don’t want to think about its function. Certainly urinals would be useful in houses with a large male population, but a urinal in a house would cause consternation among our friends and neighbors. More handrails would be useful, but they imply that the household residents are not all Olympic gymnasts. Bathrooms and the products within them seem to be culturally off limits, along with functions such as elimination, shaving of the armpits, and treating zits. So we continue to torture our bodies and flush large amounts of fresh clean water away, often to dispose of just urine, a comparatively benign fluid until mixed with the other components of sewage.
Let’s consider four categories of fit between people and products. The first is physical—the interaction of our bones, muscles, hearts, and lungs with built objects. The second is concerned with the senses—vision, touch, smell, and so on. The third is cognitive—the mind-machine interaction. Fourth, we will talk about problems resulting from system complexity. Each category demonstrates different aspects of the problem of fit, and each has its own wisdom necessary for the designer of good products. We will end the chapter by briefly considering some issues of safety and health.
Historically, the physical aspects of fit were the first to be encountered. As designers of tools, we evolved certain shapes that were compatible with the human body. Probably the first rocks to be used as tools were selected because they fit the hand well and had the proper weight to be moved by the arm. If you intend to use a sword in battle, you do not want it to slip out of the hand. If you pitch hay for a living, the diameter of the fork handle is important. Humans also learned the limits of their physical capability—if you shovel dirt all day, you learn how fast you should shovel.
Life through much of history was technically simple enough that such knowledge could evolve slowly and reside in the minds of the toolmakers and users. Unfortunately, such simplicity was not to last. As the great industrialization movements of the 18th, 19th, and 20th centuries swept the world, all manner of new and less natural human activities evolved: working bent over in dark and dusty coal mines, reacting to the speed of steam-powered machines and transportation devices, dealing with the complexity of mass manufacture. By the beginning of the 20th century, industrialization had outrun the poor human. Industrial accidents and job-related illnesses were taking a large toll. Products were being designed that took little account of the strength or flexibility of the operator and demanded tolerance of extreme temperatures, sound levels, smoke, and increasingly toxic substances.
At the time, the industrial worker attracted increasing attention. Historically, the human had defined the work. Now it was necessary to think harder about the relation between the human and the work. The main interest of management at the turn of the century was output and profit—to produce more with a given amount of labor. As a result, a whole field evolved centered upon the productivity of human workers. As factories became more complex, it became evident that the integration of people and machines was important.
A much-maligned pioneer in this field was Frederick W. Taylor, who along with colleagues such as Luther Gulick, Lyndall Urwick, Henri Fayol, and others developed the field of “scientific management.” This concept included hierarchical management with unity of command and limited span of control. In the early 1900s, the models for successful organizations were the military, the church, and large governments, so scientific management also assumed top-down authority, standardization of job design, and uniformity of behavior. Scientific management viewed people in factories as components of the overall machine and sought to design their jobs accordingly. The concepts of scientific management were extremely successful in increasing productivity, although now they seem rather inhuman. To the advocates of scientific management, the physical design of the job was critical, and once designed the worker should perform it repetitively, automatically, and incessantly. Be that as it may, scientific management was popular in its day, and much effort went into better understanding the physical relationships between humans and machines. Scientific management’s weakness lay in not considering the psychological problems associated with endless repetition of trivial acts. (Psychological consideration was to come later with the increasing influence of industrial psychology experiments beginning in the 1930s that showed the importance of such factors.)
A large boost to the understanding of physical fit came during World War II. During World War I, the number of machines was relatively small and the machines’ performance was such that control of them presented no great problem. Airplanes were unstable and often tricky to fly, but speeds were low enough that people could learn to deal with the idiosyncrasies of the machine. Tanks were not designed with the human in mind, but they were few in number, were slow in speed, and did not operate over long ranges. The situation was much different in World War II. Speeds became high enough that hard thinking was required to ensure human operators could adequately control the equipment. Missions of such machines as submarines, bombers, and armored vehicles became longer. In addition, the number of machines was enormous, making it necessary to design them in such a way that a wide variety of people could operate them. They necessarily were made more user-friendly. A P-51 took more physical effort to control than a present-day fighter does. However, the plane was a long way from a World War I fighter as far as stability, dependability, and comfort. By the end of World War II and after, standardization and logical control and display layout also were given a great amount of attention.
A great deal of effort from the military, NASA, and industry (perhaps more than the private sector) continues to be focused on further understanding the attributes, performance, and abilities of humans. When teaching my course, I would simply grab a couple of armloads of books and journals on the topic from the engineering library and take them to class. The students were consistently amazed at what information about humans is known. You want to know the average shoe size of women who are in the fifth percentile of height? No problem. A curve of motion-sickness onset versus oscillation frequency? Easy. The interesting thing to me, since I knew about this mountain of information, was that the students, who had been in the library many times, did not even know it existed.
Data exists on almost any physical aspect of the human. In addition, mockups of a product can and should be made during the design process in order to ensure an even better fit to the user. There really is no excuse for products that are not a good physical fit to the humans who will come in contact with them. However, such products exist in abundance. Think of the size and labeling of the controls on audio systems and miniaturized electronics, bathtubs when used by the elderly, and neckties.
The process of designing products for better physical fit is straightforward. The first step is an increase of awareness and the desire to better match product and user. This step requires constant vigilance in organizations. The companies most successful at this often feature a crusading leader who is dedicated to the fact that products should serve people, not make their lives more difficult. The second step is knowledge of the information and data that is available and how to produce more, if needed. The third is adequate testing with users before producing the product. All three of these steps are often neglected.
Although frequent prototyping is invaluable during the process of designing a product, it is often minimized because designers believe they can solve all potential problems in their heads, while organizations attempt to reduce costs. Testing time can also be a problem. I am always amused by watching people in stores testing mattresses by lying on them, and then immediately getting up, thinking that they have adequately judged the product. Obviously sleeping on them for at least one night is required, but if a mattress feels good on initial contact, the customer is satisfied. Designers often use the same approach. If the tool feels good to hold, they assume it will probably continue to feel good after a day or a week of use. Not necessarily! One does not detect carpal tunnel syndrome after the first 20 minutes of using a keyboard. The problem is that not only do theoretical techniques not solve all problems, but some problems will not even be recognized until an actual trial is conducted. It is naive to believe that we are smart enough to predict everything that may go wrong through thinking, analyzing, testing of components and subsystems, and computer simulation.
There are also new reasons for addressing physical fit: demographics show an aging population. No matter what exercise programs, medication, and diets are followed, people’s physical abilities and characteristics change. Just because I still fit in my 1970 sports car does not mean that I am still flexible enough to easily get into and out of it. Our expectations also change as we age. We expect more from our products. My foot does not fit on the accelerator of said sports car unless I take my right shoe off. It is a Jaguar XKE, and I bought it a number of years ago for love, not physical fit. At the time, I thought it only reasonable that it required me to take my shoe off, but I don’t want to have to do that anymore.
Another reason for addressing physical fit in design has to do with changes in lifestyle. For example, employment patterns are taking more of us out of active work and sticking us behind computers, displays, and buttons. Simultaneously we hear that we need more exercise. The process of designing workspaces, equipment, and methods of acquiring exercise that are compatible with the restrictions of urban life and the human animal is a challenging one.
And how about people with disabilities? Many such people who were formerly neglected in the design of products now must be considered. There will be more and more emphasis on this group as the population ages. For example, vision and hearing impairment will no longer be considered anomalies. Better methods of getting around will propagate. There are already places such as shopping centers in retirement areas where people traveling by automobile or on foot are at the mercy of those in souped-up golf carts and three-wheeled scooters.
The good news is that in general, people, aging or not, seem to be less forgiving of physical products that do not physically fit them than they used to be, and products are becoming more accommodating. The griping about cramped airplane seats is causing airlines to loosen up seating. The SUV became popular partly because automobiles became less compatible with the stuff people haul around in them. I revel in the easy availability of sizes that fit people of my build (6′2″, but once 6′4″, 230 pounds—did you know you are going to shrink?). It used to be most difficult to buy long shirts and coats and wide shoes, but now they are all over the place. But there is still a long way to go.
The model of reality built by our brain is based on information from our senses. They are usually grouped in the categories that follow:
Hearing sense—sound
Movement senses—vestibular (orientation), kinesthetic (body configuration)
Skin senses—pressure, heat and cold, pain
Chemical senses—taste, smell
Organic senses—state of body (hunger, sexual satiation, and so on)
Time sense—passage of time
The first five categories have specific sensors to detect the pertinent information. Sensing is accomplished by the sensors, the nervous system, and the brain acting as a unit. The process involves memory and is affected by emotional and cultural factors. If we fail to hear the phone ringing at a party, it is difficult to say whether the ear did not detect it or whether the brain merely did not attend to it. Our senses do not give us the true or complete state of the world. They have evolved over many millions of years to relay particular information in a way that is consistent with our survival, much of which has been in a hunter-gatherer outdoor existence. Our senses, however, are not particularly optimal for our present lives. Since we now are awake more at night, we would be better off if our vision extended more strongly into the infrared region. It would be nice if our sense of time were a bit sharper to be consistent with the clock-driven culture we have designed for ourselves. Perhaps we need a sense to tell us if we are being hustled.
One of people’s most common errors is to assume that their models of the world are reality, when they are actually only their personal reality. These models do not necessarily correspond to nature; they are merely made from the information processed by the brain to which the sensors are sensitive. Obviously in the case of products where the senses are key (such as food, drink, perfume, and wallpaper), designers must be extremely sophisticated about their role. However, the senses are a bit more subtle than the skeleton and the muscles and perhaps less likely to be paid adequate attention by many product designers. For example, you probably have a pretty good picture of what happens when you bend your elbow: a signal from your brain causes the appropriate muscles to contract, pulling upon tendons fastened to bones. But what happens when you smell something? Do particles from the rose blossom actually enter your nose? Probably not. If the scent is conveyed by chemicals, what chemicals correspond to rose smell? Even more, what sensors in your nose identify them, and how? Then the $64 question: How do the signals from the sensors create the sensation that you interpret as the smell of a rose?
Even those senses whose mechanism we think we understand remain a mystery when the brain is involved. We know that the eye, for example, contains a lens and an array of sensors called the retina. These sensors are specialized to recognize different wavelengths and intensities in the image. They then send the appropriate signal through the optical nerve. So far so good. But then what happens? The signals go directly to a three-pound piece of meat, out of which comes your visual reality. How on earth does it do that? I am always amused by the facetious theory that there is a homunculus (a tiny person) sitting in one’s head watching a TV set that is connected to the optical nerve.
The first five senses in the list earlier in this section have some characteristics in common. First, they have sensitivity ranges that cause them to detect certain types of intensities and frequencies of information and not others. A term that is often encountered when studying the senses is threshold, referring to the level of signal necessary before the sensor and brain recognize something is happening. Thresholds are low for traditional signals (touch—you can feel the wing of a fly falling on your cheek from a distance of one centimeter; smell—you can detect one drop of perfume diffused into the entire volume of a six-room apartment) and high for nontraditional signals (high-frequency sound—you can’t hear it at a frequency above 20,000 hertz). Second, the senses tend to be more sensitive to changing information than to a constant input. In fact, the senses typically become insensitive to constant signals over time. Finally, sensitivity to change varies from sense to sense. A change in sound pitch of 1/133 of the original level can be detected, but salinity of water must be increased by one-fifth before a taste difference can be noticed. In general, there is a correlation between sensitivity and survival value. Salinity is a general measure of environmental interest, and it is of importance to us to detect major shifts. Sight and hearing detect instantaneous events that could immediately spell danger.
Successful products take account of all such characteristics. Effective warnings to equipment operators give indications to the eye, the ear, and maybe even touch (stall warning in aircraft). Fire alarms are cleverly tuned to the frequencies to which we are most sensitive. Designers of radar displays are aware of the negative effects of long periods of scanning an unchanging picture. Humans are often assisted by electromechanical devices (camera light meters, carbon monoxide alarms) if they must detect intensities or changes in stimuli that the senses cannot discriminate. However, many products do not take adequate account of these characteristics of the senses. The signals emitted by electronic devices such as cell phones may be inaudible to many aging males who have lost their sensitivity to certain frequencies. Many visual warnings do not take account of the large percentage of males that are color-blind. There are product opportunities in compensating for these characteristics.
Let us consider hearing in a bit more detail in order to gain an appreciation for the limitations of our senses, the distortions that they place upon our perception of reality, and the challenges facing the designer of good products. Hearing is accomplished through a transducer (the ear) that converts mechanical motion to electrical signals. It has been estimated that for the detection of sound frequencies near 3,000 cycles per second, the vibrations of the eardrum may be as small as one-billionth of a centimeter. The human ear is able to handle tremendous variations in the power of sound. The loudest sound we can stand is about 100 trillion times as powerful as the weakest that we can detect. However, the ear is quite limited in frequency. The usual rule of thumb is that humans can detect signals between 20 and 20,000 cycles per second. However, most of us are not that good. As we age, we lose our ability to detect the higher frequencies.
Like the eye, the ear has evolved to handle the signals that occurred in nature in a less technological time. It is a wonderful device but does a good bit of interpreting, especially when hooked to the brain. Be impressed the next time you are barely able to understand a conversation at a party or over a bad phone connection. If your brain were not equipped with expectations, you would not be able to understand it at all. Also be impressed at how sensitive the ear is at the frequency of a baby’s cry.
The inner ear is essential to our ability to detect orientation and movement. The vestibular senses make use of detectors in the inner ear called the utricle and the semicircular canals. The utricle is basically a bowl with pebbles inside and sensors extending through the walls of the bowl that measure movements of the pebbles due to gravity and other accelerations. The semicircular canals are filled with fluid and detectors to measure the motion of the fluid under rotational accelerations. These sensors together give us a combination of orientation and acceleration. However, they cannot measure velocity and displacement directly. The only way the brain can get a sense of velocity and displacement is to integrate the outputs of these sensors, and they are not the most accurate sensors imaginable. The may also fail to deliver accurate data in nonstandard environments, such as space, where gravity does not give its customary signal.
The outputs of the sensors related to hearing are often augmented by information from the other senses, such as vision, touch, and the kinesthetic senses that tell us about the configuration of our bodies. Our senses are integrated. Our appreciation of food depends upon sight, smell, taste, and perhaps touch and sound. Head position is inherent in locating the source of a sound, and a number of our senses may be needed to ascertain whether it is our train that is moving or the one on the next track. Vision helps the inner ear determine our orientation while flying an airplane, and if the world is not visible due to weather, we add an instrument to look at.
It is poor understanding of the senses associated with the ear that causes product problems such as the cell phone the elderly cannot hear, the potential long-term damage from audio amplifiers, and the unmuffled leaf blower. It is the mind’s input of expectations based on previous contact with humans that make computer-generated voices so much more annoying than the programmers might expect. One wonders also how much the designers of automotive suspension systems and seats understand about the vestibular senses. Certain “luxury” cars are extremely efficient at inducing motion sickness, as are the rear seats in many buses.
As in the case of physical fit, sensitivity to problems having to do with the senses is necessary in order to produce high-quality products. As an example, smell is one of the most evocative senses and is particularly interesting because of its ability to result in strong feelings and evoke powerful memories. For example, Plasticine clay, the type used in kindergarten, often evokes memories of graham crackers and milk in the elderly, who were served these delicacies in kindergarten. On the other hand, the smell of decomposing animals is offensive to all of us.
It is not uncommon for real estate agents to bake a batch of cookies in a house before showing it. Architects have to worry about the combination of material used in a house producing long-lasting and unwelcome smells. Used cars are often dosed with a bit of “new car” smell. You probably have your own weak spots. One reason I keep the Jaguar that I no longer fit in is that I am a sucker for the mixed leather-and-leaking-oil smell of vintage English cars. As an extreme example, I once had a student who admitted to the problem of loving the smell of new pocket calculators. He claimed he had to hold his breath while passing the calculator counter in the university bookstore or he was likely to buy yet another one. Manufacturers of perfume and food are certainly sensitive to attractive and repellant smells and use their knowledge and skill to entice consumers to use their products. Many other manufacturers, however, are not so sensitive, even though smell could be a major attractor or detractor. I bought a rather nice-looking small rug a couple of years ago that I thought had a slight, but not bothersome, odor at the time of purchase that I was sure would disappear. The smell not only remained but became bothersome indeed, and I eventually gave the rug away.
Though we are rapidly gaining knowledge about the brain through experiments with such research approaches as functional magnetic resonance imaging (fMRI), we are still surprisingly ignorant of how it works. Twenty or thirty years ago, people liked to refer to the “cognitive revolution.” They were talking about our increasing understanding of the brain and its function. But people were at the same time talking about the information revolution—our increasing ability to process and transfer information due to developments in technology. The information revolution has been doing much better than the cognitive revolution. Such technological breakthroughs as the microprocessor and the satellite have enabled us to design products that allow access to huge amounts of information. A study at the University of California, Berkeley, by Peter Lyman and Hal Varian concluded that about 5.4 billion gigabytes of new information was stored on paper, film, magnetic discs, or optical discs in 2002.2 If this information were all converted to print, it would fill half a million libraries the size of the Library of Congress. Forty percent of the information was produced in the United States, which if converted to print would provide eight pickup loads of books for each citizen. And, of course, as the Internet grows, the information stored online will continue to explode.
Unfortunately, there is no indication that shows our brains are becoming more powerful at a rate consistent with our ability to manufacture and store information. Humans therefore have a problem. We are in danger of becoming overwhelmed by information. I could dwell on the point that the majority of this information is of low quality, being collections of data that have been neither sorted nor put in a useful form, but the sheer quantity alone means that we are becoming cognitively more burdened.
It is not surprising that our brain, although inconceivably wonderful and capable of incredibly rich function, is limited, since it is made of a large but finite number of simple cells. If we take into account the slowness of signal movement (the speed of the signal in the neuronal axon—the brain’s “information conductor”—ranges between 1 and 250 miles per hour, very slow compared to the speed of an electrical signal in a wire—186,000 miles per second); the redundancy needed to learn new operations and perform parallel processing; and the allocation of neurons to fixed circuits, we realize that the brain is, in fact, severely limited in speed, capacity, the number of things it can do at once, and many other functional ways.
The conscious mind does not like to do two things at once. The only way it can do so is either to alternate tasks or to integrate separate functions into one act. (Does this make you wonder about multitasking? You should. Research being done at the time of this writing indicates that it is vastly overrated.) The conscious mind also does not like to dwell on its own limitations. We don’t think about what our products require from the limited mind, or about the reactions of such minds to the products. In fact, we are so used to considering the mind omniscient that we find thought exercises annoying and our response to our difficulty with them surprising. (For example, as quickly as possible, think how many capital letters of the English alphabet have curved lines in them—not that fast, right?) Such exercises are frustrating, and frustration is an emotion that keeps us from apprising ourselves of our limitations. This emotion affects designers, just as the rest of us, and can get them into trouble when they design products. For example, although the human attention span is limited, designers may forget to take this into account when creating displays.
When the mind is involved, people must also consider not only individual products but also the total amount of work they ask their brains to do. I do not believe that any single electronic device is too difficult for the human mind to comprehend. However, some people seem to be baffled by them. It is probable that the rapidly increasing sum of knowledge needed to operate all electronic devices is being resisted by the mind due to its sensitivity limits. Consider present-day organizational systems, complete with voice mail, electronic mail, several types of traditional mail, telephones, smartphones, Facebook and Twitter accounts, intranets, faxes, beepers, cellular phones, satellite links, and a myriad of computer and television networks. Do you worry about increasing demands upon the attention of drivers as automobiles routinely become outfitted with car phones, map displays, and full-function computers that allow the driver to access e-mail, write memos, create spreadsheets, or just chat with friends? You should.
Fortunately, businesses have discovered that money can be made by using technology to help us manage our technology. When I recently redid our home entertainment center, which I have put together from various components over time, we ended up with five separate remote control units (audio, TV, cable box, DVD, VCR). In order to prevent cognitive breakdown, I spent money I didn’t want to spend on a Logitech universal remote.
The mind is also not as “logical” as we might think. The human brain is often compared anatomically with the brains of lower animals. Many models use this anatomical similarity to explain similar behaviors between the two groups. In 1960 a psychologist named Paul MacLean proposed the Triune Brain theory, which has had a great deal of influence as a model.3 This theory divides the brain into three mechanisms—the R (for reptile) complex, the limbic system, and the neocortex, mechanisms for reacting, feeling, and thinking. In general, lower animals engage in ceremony and hierarchy, just as we do. For example, the reptile goes home to lay its eggs, while we humans visit our hometown, even though it may have no logical attraction. Such behavior is sometimes attributed to the hindbrain, common to both reptiles and humans and operating on inherited programming—not logical at all.
The primitive midbrain is concerned with processing and switching signals from the senses and the motivation of the animal. The same is true of our midbrain, which contains the thalamus, which is the relay station for sensory information, and the hypothalamus, which is involved in behavior having to do with basic biological urges (eating, drinking, sex). Many of our basic emotions emanate from the midbrain, which does not always respond in “logical” ways: we fall in love with the wrong person or are unreasonably prejudiced against ethnic groups or homosexuals, even though we are not supposed to be. We are suspicious of strangers, even though we may tend to like people once we get to know them—an emotional response, not a logical one. It is the highly developed human cortex, or forebrain, that we credit with bringing us our logical ability, of which we are rightly proud. But we must remember we do have midbrains and hindbrains.
There are many models that seek to describe brain function. A great deal of research is being done on the topic. Many older but still interesting models are portrayed in a fascinating manner in a book entitled Maps of the Mind, by Charles Hampden-Turner.4 Most of these models speak to a large amount of unconscious function coupled with consciousness. This duality between consciousness and unconsciousness often gets companies into trouble when designing products—they design them consciously. However, when using these products, both designers and consumers employ a mind that is heavily dependent upon past experience and is quite habitual in its actions. Designers therefore necessarily exclude vast amounts of data and options when reaching conclusions, simplifying life in order to cope with complexity. They are not, however, necessarily conscious of doing so while in the process and often fail to produce products that are as compatible as they should be with humans. Designers are at the mercy of their own problem-solving habits, which probably do not match those of the potential user of the product.
As an example, let us think about computer products. Computers do large numbers of extremely simple logical actions at high rates of speed. They are therefore suited well for mathematical calculations and performing other functions where there is a one-to-one correlation between input and output. As their capability has increased, however, computers have been turned to the solving of increasingly complex problems. For some time, there has been an ongoing debate as to how capable computers will eventually become. One pole of the argument (the hard core of the artificial intelligence community) feels that computers will eventually be able to do much, if not most, of the type of problem solving that the human mind now does. In fact, the more visionary extremists feel that a “silicon mind” would be superior to a flesh-and-blood mind, in that it would not be subject to biological death and therefore could keep on learning and developing forever.
The other side of the argument is that barring unpredictable breakthroughs having to do with neural networks, parallel programming, and biological elements, computers are never going to handle the complexity and uncertainties that the human mind does. For instance, if my wife gives me a simple request, such as “Could you pick up some fruit on your way home?” I fill in a tremendous amount of information from my experience. I know that the proper way to respond is to go buy some fruit rather than merely answering “yes.” I know that the pick in the sentence is different from a guitar pick or the pick I use to help dig holes in my yard. I know that I have to go to a grocery store and the incredible myriad of details having to do with driving there, finding the fruit, deciding what is the best deal, getting it through the checkout counter, and so on. I can also handle the many unprecedented situations I will encounter driving to and from the store and maneuvering my shopping cart through people of all ages, sexes, and moods. It will be quite a while before a computer can do all of that.
There is a tremendous amount of work going on in the area of computer problem solving, and our society is certainly giving the computer the benefit of the doubt. We are adopting it, learning to use it, and even learning to act more like computers in order to deal with it—but herein lies the problem. Manuals and the Internet contain vast quantities of information often presented in the jargon and framework of computer designers. There are some people who love to decipher this information. However, this is not true of most people, and computer inputs and outputs must become much more sensitive to the way humans think.
The ancient screen saver program After Dark was a good example of a program that was compatible with the human mind. Each routine had similar controls, and each could be viewed and adjusted rapidly before being activated. The Internet and the associated World Wide Web, as I write this book, are not compatible with most human minds. They contain too much redundancy, too much information of little value, and inadequate sorting. Like most of my friends, I find them useful as intellectual entertainment and to find specific things, but otherwise overwhelming. As an example, I just typed “product quality” into Google and received 147,000,000 items—a bit too many to go through. One of the problems may be that it only took the browser program 0.15 seconds to find all of them. Perhaps even computers cannot pay much attention to quality in that amount of time.
Humans have arrived at a stage in our history when we have learned to make extremely complex technological systems. Of course, we have been creating complex systems for a long time, as shown by the Roman aqueducts, the British canals, and the American railways. These earlier systems, however, consisted of rather independent components of a reasonably simple nature. It was possible, for instance, for a human operator to understand a locomotive sufficiently to not only control it but also respond to abnormal situations. Now consider present systems such as the space shuttle, modern passenger aircraft, and nuclear reactors: here, we enter into serious complexity. How about nuclear weapons systems and reengineered ecologies? Maybe we are in over our heads. In an excellent, and controversial, book called Normal Accidents by Charles Perrow, the author suggests that as long as humans are involved, there will be accidents in systems above a certain complexity. He states that if the result of an accident could be unacceptable to society as a whole, as in the case of a strategic nuclear missile system, perhaps we had better lay off. In a system of great complexity where the costs of an accident seem to be tolerable to the society, such as the U.S. aircraft control system, he says go ahead.5
Complexity relates directly with the topics discussed previously in this chapter. Obviously, it is easier for humans to interact with a complicated system that is consistent with their physical, sensory, and cognitive capabilities. However, at some point, complexity itself becomes an issue. As an example of the type of problems we can get ourselves into, let us briefly consider the previously mentioned accident at the Three Mile Island nuclear energy plant in 1979. This accident received a great amount of attention and was thoroughly investigated. In fact, the report of the Nuclear Regulatory Commission’s independent special inquiry group directed by Mitchell Rogovin on the Three Mile Island accident is a fascinating (if long) read.6 It is a superb portrait of the type of difficulties that can overwhelm an extremely complex man-machine system.
In the case of the Three Mile Island accident, there were no fatalities or detectable radiation-induced illnesses. But the incident so shook up the nation that it caused permanent changes in the way nuclear reactors are designed and operated and to the public’s attitude toward nuclear energy. In the view of some, this incident destroyed the nuclear energy business in the United States, although it is showing signs of resurgence at the time of this writing. The accident was due partly to equipment failures and partly to operator errors. For example, when a relief valve failed to properly perform its function (decreasing the pressure in the nuclear portion of the system), the operators didn’t know what to do—they had never been in such a situation. In a complicated system such as a power plant, it is almost impossible to train operators by simulating all possible malfunctions—the combinations of possible component failures is simply too great. The information provided to the operators was incomplete and confusing.
As a result of this accident, many changes were made not only in the hardware of nuclear power plants but also in the training of operators. This accident in a complex system was due not only to inadequate instrumentation but also to incorrect decisions made by operators who had not been adequately trained in abnormal reactor operation and who were both overwhelmed by misleading information and deprived of the information they needed. The challenge to designers of complex systems is to prevent such accidents through foresight.
The Three Mile Island operators simply did not understand the system well enough to do the proper things when it was in an abnormal state. This situation is not difficult to understand. We are all taught, for example, to operate automobiles in a normal operating state. Most driver instruction courses do not include skid pads, blowouts, and oncoming cars veering into our lanes—but we might be better off if they did. One of my friends who was a highway patrolman noted that in freeway accidents, few drivers would leave their lane. As a result, as is often the case in crowded conditions and bad weather, dozens of cars would run into each other. My friend claimed that people simply held the wheel steady and slammed on their brakes, ignoring opportunities to steer onto shoulders or into vacant lanes. He argued, convincingly, that drivers should be trained for the unusual. The same argument is made for private plane instruction. It is possible to acquire a private license without ever having been in a spin, but although modern private planes are much less likely to enter spins, not all private planes are modern, and spins continue to be a factor in small plane accidents, so it would seem to make sense that all licensed pilots be prepared to react to them.
Complexity must be taken into account in preparing human operators for emergencies. Designers of such systems as a nuclear reactor plant, air traffic control system, or refinery often understand the systems well enough that they might be able to respond to abnormal or emergency situations, but the systems’ operators may not. Either highly sophisticated simulators must be used in training operators, people with more thorough knowledge must be on hand at all times, or systems must be designed to be compatible with the limits of the human operators. Controls, instruments, manuals, and alarms must be more carefully designed as complexity increases. Redundancy must be included, and computer–human control interactions must be considered. It may be acceptable for your computer to simply show you a warning on the screen if something goes wrong, but that is definitely not adequate feedback for the operator of a nuclear reactor.
Individual consumers do not buy nuclear reactors or refineries, but we sometimes forget that we are components of an extraordinary complex system, consisting of ourselves, the infrastructure that we have built upon our planet, and the planet itself, and that we have to operate the components of that system that affect our lives. It isn’t that we have difficulty handling a few components of this system; it is the totality that is becoming a significant burden, and a major contribution is the rapidly expanding number and increasingly complex character of products.
Producers and consumers alike should worry about this situation. Customers seem to be eager for innovative products with new and different features and more capability. That’s progress, right? I think not if it is carried to the point that individuals and groups of humans spend too much of their time and effort trying to remember how to operate and maintain these products and too much money hiring professional expertise to help. Technology is supposed to help us live a satisfying life, not consume our time and energy in its operation and upkeep. I am in favor of diversification of products, but also standardization of the controls and displays that allow us to operate them. Strangely, the military seems to handle this job better than producers of products for civilians do, perhaps because in battle you don’t want to spend time figuring out how to start a vehicle or load a rifle.
At one time, when products were simpler and we were in love more with manufacturing and less with innovation and high technology, there was more standardization among products. As an obvious example, home entertainment systems consisted of radios and record players, the former equipped with a volume dial (on the left) and a tuning dial (on the right), with the on-off switch sometimes integrated into the volume dial (all the way counterclockwise). Sophisticated record players originally had their on-off switch integrated into the arm and later a speed selector and automatic changers, but the change was slow and new features tended to be the same across brands. The television was a separate device. These former systems contained nothing like the number and variation of controls and displays in contemporary systems. The same can be said for kitchen, yard, and laundry appliances, as well as mousetraps, headache remedies, baby strollers, and a myriad of other products of industry. An example of what has occurred can be seen with cars. There was a time when you could climb into any car, roll down the window, and drive it away. No longer.
I rent cars quite often and am tired of starting the windshield wipers instead of the turn signal; experimenting to find out how to turn on the lights; groping to find the hood, parking brake, or trunk release and then experimenting to see if I have the right one (usually not); and trying to figure out the cabin temperature and navigating systems while entering heavy traffic leaving the airport in a strange city. Not only is this frustrating, it is unsafe!
The idea of fighting the complexity of our lives by paying more attention to standardization as well as ease of operation and maintenance of industrial products is perhaps contrary to our ethic of creativity, innovation, and progress through advances in technology. But the burden of interacting with these products is an increasing one, especially for users who don’t keep their manuals and handbooks with librarian skill and do not want, for example, to take their laptops out to the driveway to look up information with their greasy hands, while working on the car. There must be a market opportunity here somewhere. We all believe in products being user-friendly, but as individual products become so, our growing panoply of products seems to require increasing numbers of neurons to deal with it.
Rightly or wrongly, the United States and other developed countries have become increasingly obsessed with safety. This obsession has been explained by a number of arguments, ranging from lack of sensitivity in the past through a desire for immortality to (again) too many lawyers. Our history is somewhat bleak, considering such topics as black lung disease, radiation poisoning, and industrial accidents. We are now spending much more effort protecting ourselves from ourselves.
I recently set out to acquire some climbing equipment, since I had a large dead tree in my yard that had to come down. I finally had to order it over the Internet, because I simply could not find any in rental or retail stores in my area. The reason most often given to me was unwillingness to be involved in any liability for people like me operating well beyond our skill level or physical ability 75 feet above the ground. The conservative stand would be that it is my life, and I should be allowed to risk it if I want. The liberal stand would be that society should protect me against my poor judgment. The intermediate stand is that I should not be allowed to act in a way that would incur the enormous expense to society that would result from my becoming a “vegetable” by falling and not dying (also the argument for motorcycle helmets).
Industrial safety has been an issue for years, beginning in the United States when rapid growth in the textile industry (due to the embargo of the War of 1812) resulted in the establishment of insurance companies that inspected industrial properties and suggested methods of decreasing risk so their policyholders might qualify for low rates. In 1970 the Occupational Safety and Health Act was signed into law, establishing the OSHA administration and the National Institute for Occupational Safety and Health (NIOSH), which set and enforce strict rules for industrial operation.
Consumer safety has a similarly long history. The regulation having to do with food and drugs is a good example. In 1820, eleven physicians met in Washington, D.C., to establish the U.S. Pharmacopeia, the first list of standard drugs for the United States. In 1848, the Drug Importation Act was passed by Congress in an attempt to control the importation of adulterated and unsafe drugs. In 1906 the original Food and Drug Act was signed into law. The emphasis on consumer safety has increased markedly in the past 25 years through the efforts of activists such as Ralph Nader and Bess Myerson, institutions such as the Department of Consumer Safety, and publications such as Consumer Reports. Large numbers of products have been taken off the market, ranging from toys that can shed parts children could choke on, to automobiles that ignite in accidents.
Designing products for safety is a most challenging task, since people seem to be at their most brilliant in devising ways to hurt themselves. One of my first engineering jobs was to devise a safety feature for a machine involved in the production of an aluminum part. The difficulty had to do with the cutter speed, which caused it to be almost invisible to the human eye. The noise level in the plant and the repetitiveness and speed of the operation also caused the operator to become careless during the day. The machine cycle required the operator to load the part and then press a button, which caused the part to be fed through the cutter.
My first approach, of course, was to put a guard around the cutter. But the operators would remove the guard, claiming that it inhibited the job. I then redesigned the machine so that the cutter was buried deep within it. However, another accident occurred when an operator was attempting to remove a snarl of chips and accidentally hit the button. The third attempt was to have two buttons, both of which had to be pushed to cycle the machine, thus making sure that both of the operator’s hands were on buttons rather than close to the cutter. Operators circumvented this solution by placing their lunch or other objects on one of the buttons. Inverting the button unit to prevent this resulted in one button being pushed by a knee, and so on. I finally made a big step toward helping the situation by simply requiring holes in the cutter so that it made more noise. But this was before OSHA sound regulation limits.
This ability to find danger can be clearly seen among children. My kids would constantly fascinate me by making weapons from the innocuous, tipping over the permanent, and eating the unthinkable. But we grown-ups are certainly not above reproach. Think of your own life. How many things do you do that products should not allow you to? Here are a few of my own:
• I always remove all guards from power tools. I grew up with open blades on circular saws and naked sanding belts and find guards annoying. Guards should be improved to not inhibit usage and then be made an integral part of the machines.
• I am attracted to strong solvents, such as acetone, get them all over my skin, and inhale their vapors. I welcome efforts toward potent but less dangerous chemicals, such as the continuously improving water-based paints, but I wish they would improve more rapidly.
• I ride my bicycle far too fast for its braking ability. Mountain bikes have good brakes—why not put decent brakes on street bicycles?
There are more, but I just include these to encourage you to think a bit. I highlight them with no shame, because I think I am as careful as most people. And like most people, I am not as careful as I think I am. The key to creating safe products is to keep that in mind: it is a difficult task. I find myself insulted by many of the features that are built into products to prevent me from harming myself. But on the other hand, there is great diversity in the population. Probably (hopefully) these features save some people from harm. As in the case of designing products that are consistent with physical, sensory, and cognitive limitations, the key is conscious effort to be sensitive to the human animal. Tremendous creativity and user testing is necessary to predict the possible misuses and subtle negative effects of the products of industry.
Chapter 4 Thought Problem
Here we go again. But for a little variety, this time you may choose six products—one that fits you extraordinarily well and one that fits you poorly from first a physical, then a sensory, and finally a cognitive perspective. Why are they good or bad fits to you? If you own them and they do not fit you, why do you put up with them? Again, if they do not fit you well, how could they be improved to fit you better, and why do you think these improvements have not been made?