Human Factors in Reliable Design
Human Factors Engineering
Human factors engineering (HFE) is a specialty engineering discipline that focuses on designing products and systems with the user in mind. The goal of HFE is to maximize the ability of a person or a crew to properly operate and maintain a product or a system by eliminating design-induced impediments and errors and to improve system reliability. It is important that the design engineer recognize the limitations as well as the capabilities and preferences of the human being in the system. A product that has been designed from this perspective will offer better customer satisfaction, require less training to operate and maintain, and will produce fewer errors and failures.
Human factors engineering is defined by MIL-STD-1908 [1] as “the application of knowledge about human capabilities and limitations to system or equipment design and development to achieve efficient, effective, and safe system performance at minimum cost and manpower, skill, and training demands. Human engineering assures that the system or equipment design, required human tasks, and work environment are compatible with the sensory, perceptual, mental, and physical attributes of the personnel who will operate, maintain, control and support it.”
Human factors engineering as we know it today has its origin in World War II due to military needs to design and operate aircraft safely. Prior to that time, human beings were typically screened to fit the job or equipment rather than having the equipment designed with humans in mind. Although human factors engineering grew initially in the defense and aerospace industries, it has since spread to all industries, including nuclear, space, health care, transportation, and even furniture design.
A Design Engineer's Interest in Human Factors
Since all products or systems are a collection of components that interact with each other and the external environment to achieve a common goal, the designer must be concerned about all the components and their interactions. Some of these components include humans. A human may be part of a system, a user of the system, or a controller of the system. Failures in products or systems are often blamed on human error. Quite often, human error is a symptom of something wrong with the design of the product or system. This may indicate a lack of consideration of the human aspect during the design phase of a product. Ideally, we want designers to design out all possibility of human errors. This is achievable with human-centered design practices.
Additional concerns with modern systems are the fact that the complexity of systems is increasing constantly, the technology is growing exponentially, and products and systems are becoming more difficult for people to use and understand. This makes it even more critical that humans be considered in product design. The reliability, safety, and usability of products and systems will be enhanced substantially if they are designed with people in mind. Safe and efficient operation of products and systems depends on properly designed and engineered interactions between the human and the machine.
Human-Centered Design
Human-centered design encompasses a wide variety of concerns. As the name implies, human-centered design places people at the center of design considerations rather than making people conform to the design. The design must accommodate humans' physical characteristics and mental processes such as perception and cognition. It must also take into account the environment in which a person must operate and the characteristics of the equipment, product, or system. The user also operates within an organizational framework that must also be considered. Other factors that influence human behavior, and should therefore be considered in the design process, include the technology being used, the management systems that are in place; and the procedures and processes under which the user will operate.
Role of Human Factors in Design
The product design should minimize human error and maximize human performance. Design considerations must include human capabilities, human limitations, human performance, usability, human error, stress, and the operational environment. Some typical topics that should be taken into account during the design process are shown in Table 1. This table is not meant to be all-inclusive; it is provided to highlight some major topics and to stimulate thought about what should be considered in a design.
Table 1 Considerations in Human-Centered Design
With products and systems becoming more complex, organizations must be committed to human-centered design. They must fully integrate systems engineering with all types of specialty engineering, including human factors engineering; must develop system requirements to include user requirements; must test the product being developed with real users, assess the usability, and fix any shortcomings identified; and must use the tools and techniques that will facilitate this integrated design approach.
Hardware
Designers must consider the human being during the design process as related to hardware. Typical hardware-related considerations include weight, layout, access, anthropometrics, and ergonomics. Many examples of bad hardware design can easily be found in everyday products:
- Remember the Ford Pinto which required that the engine be dropped in order to change the spark plugs?
- Have you ever rented a car, pulled into a gas station to refuel it, and didn't know where the gas cap was? It could be on the right side, left side, or under the license plate. Inevitably, it is always on the side opposite the one nearest the pump when you pull in. Why aren't they always on the same side? Why aren't they always on the driver's side, as that's probably who will fill up the car?.
- Have you ever looked at a stove and wondered which knob controls which burner?
All these design flaws could have been eliminated if the designers had taken into account that essentially all products and systems have users who need to be considered and the design needs to accommodate them.
Software
Similarly, software design should consider the user. How many times have you installed a new computer program, opened it up, and discovered that you have no idea how to do anything with it—the icons are indecipherable, the order in which things need to be done is a puzzle, the color scheme makes the fonts unreadable, and no useful instructions are provided. What about Microsoft Windows' nonintuitive design … Why would anyone design a system where one must go to the start button to shut off the machine?
Traditionally, human factors considerations in the design process have been focused on hardware aspects of the product or system. In the last couple of decades, more and more products have become more software intensive. The proportion of software to hardware in products has been increasing steadily. This trend has made development and change more rapid, and the greater use of software imparts a greater risk in both operations and maintenance. Software requires different skills; it makes diagnosis more challenging and generally imposes new considerations for the designer to think about during the design process.
Usability is a term used to describe the ease with which a person can employ a product or system. ISO 9241 [2] defines usability as the “extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use.” Although the term can be applied to both hardware and software aspects, it is most often used in relation to software. As complex computer systems find their way into our everyday life, usability has become more popular and more widely utilized in recent years. Designers have seen the benefits of developing products with a user orientation. By understanding the interaction between a user and a product, the designer can produce a better, more widely accepted product. Desirable functionality or design flaws may be identified that may not have been obvious if human factors have not been considered. Implementing this human-centered design paradigm, the intended users of the product are kept in mind at all times. Maybe, then, the user won't have to go to the start button to stop the machine!
As with all good systems engineering, the most important part of the process is the up-front definition of requirements. It is critically important that complete user-interface requirements be identified early in the development process. There are numerous guidelines and style guides for user interfaces. An example of a guideline is ISO 9241, one of a series of guidelines for various aspects of computer–user interfaces. These provide a starting place for user-interface requirements generation, but they must be customized and tailored to fit the application.
Also important to implementing user interfaces successfully is design evaluation. This can be done using mock-ups or prototyping the user interface and testing with actual users. Requirements and implementations can then be adjusted in the design process when it is still cost-effective to make changes. Continuing user evaluation as the design evolves will ensure the best usability of the end product.
Human factors specialists do more than design friendly icons. They bring two important types of knowledge to bear on systems development: (1) human abilities and limitations, and (2) empirical methods for collecting and interpreting data from people. They define criteria for ease of use, ease of learning, and user acceptance in measurable terms. New technology demands much thought about the role of the tool. The distress and aversion that many people manifest toward computerization is perfectly rational in the presence of ill-conceived design. Too many of our systems confuse the operator.
One example of a human capability is time perception, which can be shown to affect the functional requirements of a software design. Table 2 shows time perception across various tasks and media. Transaction interactions should be without a perceived wait, and the standard deviation of all transactions should be less than 50% of the mean.
Table 2 Perceptions Across Media
Human–Machine Interface
The interface between a person and a machine is of prime importance. The designer must be concerned with any and all parts of the product where a person must interact with the equipment. The human–machine interface becomes a critical component to be considered. This interface is defined as the plane of interaction between the person and the machine. It is across this plane that information and energy flow. The information is transferred across this interface from the machine to the person via displays and from the person to the machine via the controls. Therefore, displays and controls are of major importance in product or system design. Effective displays help to determine the proper action needed. Ineffective displays and controls contribute to errors that may lead to accidents.
The designer must also be concerned with the proper allocation of functions between person and machine, user and equipment. Machines are consistent; people are flexible. Machines are more capable of performing repetitive and physically demanding functions; people are more capable of performing functions that require reasoning. These are of prime significance when allocating functions to the machine or the person. This allocation of functions must ensure that the tasks assigned to each take into account what they do best, what their capabilities are, and what limitations they each have. Trade-offs between human and machine must be made regarding speed, memory, complex activities, reasoning, overload, and so on. In the next three sections we describe some of these considerations in greater detail.
Staff Requirements
The staff required to operate and maintain the system must be a design consideration. How many and what type of people will be needed to operate and maintain the system or product? Are properly qualified people available? Will training be required? If so, how much? Will these types of people be available in the future to support the entire life cycle of the system? Often, there are trade-offs that can be made during the design process that can reduce the number of people needed or the amount of training that will be necessary. For example, if the graphical user interface is intuitive to use, the time for the operator to learn to use the system can be greatly reduced.
Workload
The workload the system or product imposes on a user is a related concern. The tasks that a person must perform must be delineated. An assessment must be made of the amount of effort that each task will take both physically and intellectually. A matchup of the capabilities of the user to the tasks at hand must be ensured; otherwise, the user will become quickly dissatisfied with the product, or worse, will make errors as a result of task overload, which could lead to substantial undesirable consequences.
Personnel Selection and Training
Another important factor to consider during system design is who will be needed to operate and maintain the system. The proper people with sufficient skills and knowledge must be selected for the best fit to operate and/or maintain a system. Once selected, these people must be properly trained to do the job functions that have been allocated to them as a result of the system design process. Again, good product design can help to reduce these demands on the human operator or maintainer.
Human Factors Analysis Process
Almost any technique used in system analysis can be applied to address the human element. However, there are numerous human factors–specific analysis techniques from which a designer can chose.
Purpose of Human Factors Analysis
The overarching purpose of human factors analyses is to develop a better, usable, and safe product or system. Various human factors analyses are conducted at different times in the development process and for different reasons. Analysis of human factors assists in the development of requirements, which is a critical step in the design process. As the product or system evolves, different analyses are conducted to define the human role in the system, to ensure that the human needs and limitations are being considered, to determine the usability of the product, and to confirm the ultimate safety of the system. Other analyses can be conducted to help guarantee customer acceptance of the product or system.
While the human factors engineer may be the lead for conducting the analyses, it should be a team effort. The team may vary depending on the stage of development or the particular analysis being conducted, but the results will always be better if it is a joint effort by a cross-functional team. Participants in the human factors analyses will always include the design engineer. The team may include other specialty engineers, such as system safety, reliability, software, and manufacturing engineers. Often, team participants may include management, marketing, sales, and service personnel. The team should be tailored to enhance the particular analysis being conducted.
Methods of Human Factors Analysis
As our products and systems have become more complex, it has become imperative that the old approach—either totally ignoring human considerations or making “educated guesses” based on intuition of how best to accommodate the human—be replaced by systematic analytical techniques to better match the human being and the machine. Although it is beyond the scope of this book to elaborate on all the various human factors analyses, Table 3 presents a sampling of the analyses available to a design team. The reader can find more detailed coverage of these and many other techniques in the books of Raheja and Allocco [3] and Booher [4].
Table 3 Human Factors Analysis Tools
Human Factors and Risk
Risk is inherent in all systems, and people add an additional dimension to the risk concern. Designers must be aware of both the risk posed by humans and the risk imposed on humans.
Risk-Based Approach to Human Systems Integration
The goal of all good systems engineering design is to reduce risk throughout the development cycle. This is accomplished by applying all relevant disciplines. In the past, however, the risks associated with human systems integration have often been ignored. Engineering risks are noticed at various times during the development due to implementation problems or cost overruns. Risks associated with human systems integration, however, are usually noticed only after a product or system is delivered to the customer. These end-state problems may lead to customer dissatisfaction and rejection of the product, due to it being too difficult or inefficient to use or, worse, they may lead to human error in the use of the product, which could have catastrophic consequences.
These operating risks can be traced back to failure to properly integrate the human needs, capabilities, and limitations at an early stage of the design process. Like all risk-reduction efforts, risk reduction in the area of human systems integration must be started early and continue throughout development of a product or system. This will ensure that requirements based on human factors are incorporated. It will allow design trade-offs to consider use of the product. This will produce a high level of confidence that the product or system will be accepted and usable by the user.
As has been emphasized throughout the book, the key to design success rests in the development of requirements. Requirements involving human factors are no different. A thorough and complete effort to specify human factors requirements in the earliest stages of product or system development will reap large rewards later. During development, other approaches to mitigating risk in the area of human systems integration include using task analysis to refine the requirements, and conducting trade studies, prototyping, simulations, and user evaluations.
Human Error
It is easy to blame accidents and failures on human error, but human error should be viewed as a symptom that something is wrong within the system. Generally speaking, humans try to do a good job. When accidents or failures occur, they are doing what makes sense to them for the circumstances at hand. So, to understand human error, one must understand what makes sense to people. What “reasonable” things are they doing, or going to do, given the complexities, dilemmas, trade-offs, and uncertainties surrounding them? The designer must make the product or system make sense to users or they will be dissatisfied with it, failures will occur, or they will have accidents using it.
Designers often think that adding more technology can solve all human error problems. Quite often, however, adding more technology does not remove the potential for human error. It changes it and may cause new problems. Make sure that you know whether you are really solving the problem of human error by adding technology or are merely causing different problems. Take the simple addition of a warning light. What is the light for? How is the user supposed to respond to it? How does the user make it go away? If it lit up before and nothing bad happened, why should the user respond to it now? What if the light fails just as it is needed?
Many error-producing conditions that may cause a failure or an accident can be added to a product or system inadvertently. The designer must remain cognizant of the many things that can lead to human error, including environmental factors such as heat, noise, and lighting; confusing controls, inadequate labels, and poor training; difficult-to-understand manuals or procedures; fatigue; boredom; and stress. The goal is to eliminate those things that can contribute to human error.
Types of Human Error
Human error affects system reliability. There are many ways in which people can make errors. They can commit errors in calculations; they can choose the wrong data; they can produce products with poor-quality workmanship; they can use the wrong material; they can make poor judgments; and they can miscommunicate (just to name a few). Borrowing from Dhillon [6], we categorize human error using the design life-cycle perspective:
Design Errors
These types of human errors are caused by inadequate design and design processes. They can be caused by misallocation of functions between a person and a machine, by not recognizing human needs and limitations, or by poorly designed human–machine interfaces.
Operator Errors
These errors are due to mistakes made by the operator or user of the equipment design, and to the conditions that lead up to the error being made. Operator errors may be caused by improper procedures, overly complex tasks, unqualified personnel, inadequately trained users, lack of attention to detail, or nonideal working conditions or environment.
Assembly Errors
These errors are made by humans during the assembly process. These types of errors may be caused by poor work layout design, distracting environment (improper lighting, noise level, inadequate ventilation, and other stress-inducing factors), poor documentation or procedures, or poor communication.
Inspection Errors
These errors are caused by inspections being less than 100% accurate. Causes may include poor inspection procedures, poor training of inspectors, or a design being difficult to inspect.
Maintenance Errors
These are errors made by maintenance personnel or the owner after a product is placed in use. These errors may be caused by improper calibration, failure to lubricate, improper adjustment, inadequate maintenance procedures, or designs that make maintenance difficult or impossible.
Installation Errors
These errors can occur because of poor instructions or documentation, failure to follow the manufacturer's instructions, or inadequate training of the installer.
Handling Errors
These errors occur during storage, handling, or transportation. They can be the result of inadequate material-handling equipment, improper storage conditions, improper mode of transportation, or inadequate specification by the manufacturer of the proper handling, storage, and transportation requirements.
When most people think of human error, the natural tendency is to think of an error by the operator or user. However, as can be seen from the categorization above, a major cause of human error that must be considered by the design engineer is design error. One must realize that during the design process, design flaws may be introduced. During the manufacturing process, assembly errors may occur. During quality inspections, product shortcomings may not be found due to human error. During subsequent maintenance and handling by the user, errors can occur. So the designer must be aware of all these types of defects and ensure that they are considered, and hopefully eliminated, during the design process.
Mitigation of Human Error
Mitigating risks due to human factors should begin early in the design process. Although it is always most desirable to completely eliminate conditions that lead to human errors, it is inevitable that human errors will occur. Therefore, error containment should be considered in the design. As technology has evolved, we have become more and more interconnected and the consequences of a failure, whether human-caused or not, can grow to enormous proportions due to this added complexity and interdependencies.
As an example, consider the August 14, 2003 failure of the power grid in the northeastern and midwestern United States and Ontario, Canada. The failure began when a 345-kV power line made contact with a tree in Ohio. Once the failure began, a chain reaction of events occurred due to numerous human errors and system failures. The failure propagated into the most widespread power failure in history, affecting 10 million people in Canada and 45 million people in eight U.S. states. The power outage shut down power generation, water supplies, transportation (including rail, air, and trucking), oil refineries, industry, and communications. The failure was also blamed for at least 11 deaths.
As products and systems continue to become more complex and interwoven into our culture, designers must not only ensure that human–machine interfaces are understood and usable, but must also consider the potentially far-reaching consequences that a failure might have, and design products and systems in a way that will minimize these consequences.
Design for Error Tolerance
Error-tolerant systems and interfaces are design features worthy of consideration. Error-tolerant systems minimize the effects of human error. Error tolerance capabilities added to a system improve system reliability. Human error is frequently blamed for accidents, especially high-consequence accidents. It is not uncommon to hear 60 to 90% of accidents attributed to human error. While we should be eliminating opportunities for human error to occur, we should also be designing products to be error tolerant. Designers need to consider the consequences of failures of their designs. Errors lead to consequences; consequences should be minimized or eliminated. This is the essence of error-tolerant design. A design that tolerates errors avoids the consequences. By providing feedback to the user on both current and future consequences, compensating for errors, and providing a system of intxelligent error monitoring, a design can be made to be more error tolerant. One should note that the emphasis is on “intelligent” monitoring and feedback. The system should not just provide an indecipherable error message, such as “ERROR #404.” This is the type of error message so often seen these days by the average computer user, who has no idea what the problem is or how to fix it. The system should provide a more useful message that describes the problem and provides useful information to the user about the error and what should be done to resolve it.
Checklists
Checklists can be an effective tool to help a designer and/or a user. There are two main types of checklists. One is to provide the designer with a checklist to follow during the design and testing of a product or system. The second is to provide the user with a checklist to follow when using the product or system so that nothing will be omitted.
A human factors design checklist is typically a long list of design parameters that should be considered during the product design process. These lists can be based on previous experience, lessons learned, and/or some published design guide. For example, a checklist derived from MIL-STD-1472 [7] can be used to help a design engineer ensure that a design will be usable by people for its intended purpose. It should be realized that checklists have limitations. It is impossible for a checklist to cover all variables and combinations of conditions for all designs. Despite that shortcoming, checklists provide guidance to the designer for things that need to be considered. They can also serve as tools for test engineers to verify that the product or system has been designed and produced with the user in mind.
Checklists are created for a user to ensure that the product or system is operated correctly, that all procedures are followed, and that uncertainty in operation is avoided. A well-known example of a checklist is the preflight checklist, used by pilots prior to take off.
Testing to Validate Human Factors in Design
Human factors validation is just as important as the development of adequate requirements for product or system specification. It is important to test the product or system against each human factors requirement and to verify that the requirement has been met adequately. Human performance requirements should be validated in system test plans and demonstrated in usability tests, and the results addressed in test reports. The product or system should be tested by representative users to verify that it functions as planned and can be operated properly and safely by the intended user.
References
[1] Definitions of Human Factors Terms, MIL-STD-1908, U.S. Department of Defense,Washington, DC, Aug. 1999.
[2] Ergonomics of Human System Interaction, ISO 9241–11, International Organization for Standardization, Geneva, 1998.
[3] Raheja, D. G., and Allocco, M., Assurance Technologies Principles and Practices: A Product,Process, and System Safety Perspective, 2nd ed., Wiley, Hoboken, NJ, 2006.
[4] Booher, H. R., Handbook ofHuman Systems Integration, Wiley, Hoboken, NJ, 2003.
[5] U.S. Army, MANPRINT in Acquisition: A Handbook, U.S. Department of Defense, Washington, DC, Apr. 2000.
[6] Dhillon, B. S., Design Reliability: Fundamentals and Applications, CRC Press, Boca Raton, FL, 1999.
[7] Human Engineering, MIL-STD-1474F, U.S. Department of Defense, Washington, DC, Aug. 1999.