HUMAN-CENTERED PRODUCT PLANNING AND DESIGN:DESIGN METHODOLOGY

DESIGN METHODOLOGY

Concepts such as user-centered design, user-friendly systems, and ergonomically designed systems have been around for quite some time. Virtually everybody endorses these ideas, but very few people know what to do in order to realize the potential of these concepts. What is needed, and what this chapter presents, is a methodological framework within which human-centered design objectives can be systematically and naturally pursued.

Design and Measurement

What do successful products and systems have in common? The fact that people buy and use them is certainly a common attribute. However, sales are not a very useful measure for designers. In particular, using lack of sales as a way to uncover poor design choices is akin to using airplane crashes as a method of identifying design flaws—this method works, but the feedback provided is a bit late.

The question, therefore, is one of determining what can be measured early that is indicative of subsequent poor sales. In other words, what can be measured early to find out if the product or system is unlikely to fly? If this can be done early, it should be possible to change the characteristics of the product or system so as to avoid the predicted failure.

This section focuses on the issues that must be addressed and resolved for the design of a new product or system to be successful. Seven fundamental measurement issues are discussed and a framework for systematically addressing these issues is presented. This framework provides the struc- ture within which the remainder of this chapter is organized and presented.

Measurement Issues

Figure 1 presents seven measurement issues that underlie successful design (Rouse 1987). The ‘‘nat- ural’’ ordering of these issues depends on one’s perspective. From a nuts-and-bolts engineering point of view, one might first worry about testing (i.e., getting the system to work), and save issues such as viability until much later. In contrast, most stakeholders are usually first concerned with viability and only worry about issues such as testing if problems emerge.

A central element of human-centered design is that designers should address the issues in Figure 1 in the same order that stakeholders address these issues. Thus, the last concern is, ‘‘Does it run?’’ The first concern is, ‘‘What matters?,’’ or, ‘‘What constitutes benefits and costs?’’

Viability

Are the benefits of system use sufficiently greater than the costs? While this question cannot be answered empirically prior to having a design, one can determine how the question is likely to be answered. How do stakeholders characterize benefits? Are they looking for speed, throughput, an easier job, or appealing surroundings? What influences their perceptions of these characteristics? How do stakeholders characterize costs? Is it simply purchase price? Or do costs include the costs of maintenance and, perhaps, training? Are all the costs monetary?

Acceptability

Do organizations / individuals use the system? This is another question that cannot be answered de- finitively prior to having the results of design. However, one can determine in advance the factors that are likely to influence the answer. Most of these factors relate to the extent to which a product or system fits into an organization’s philosophy, technology, and so on.

Validity

Does the product or system solve the problem? This, of course, leads to the question, what is the problem? How would you know if the problem was solved, or not solved? The nature of this question was discussed earlier in this chapter.

Evaluation

Does the system meet requirements? Formulation of the design problem should result in specification of requirements that must be satisfied for a design solution to be successful. Examples include speed, accuracy, throughput, and manufacturing costs.

Demonstration

How do observers react to the system? It is very useful to get the reactions of potential stakeholders long before the product or system is ready for evaluation. It is important, however, to pursue dem- onstration in a way that does not create a negative first impression.

Verification

Is the system put together as planned? This question can be contrasted with a paraphrase of the validation question—is the plan any good? Thus, verification is the process of determining that the system was built as intended, but it does not include the process of assessing whether or not it is a good design.

Testing

Does the system run, compute, and so on? This is a standard engineering question. It involves issues of physical measurement and instrumentation for hardware, and runtime inspection and debugging tools for software.

A Framework for Measurement

The discussion thus far has emphasized the diversity of measurement issues from the perspectives of both designers and stakeholders. If each of these issues were pursued independently, as if they were ends in themselves, the costs of measurement would be untenable. Yet each issue is important and should not be neglected.

What is needed, therefore, is an overall approach to measurement that balances the allocation of resources among the issues of concern at each stage of design. Such an approach should also integrate intermediate measurement results in a way that provides maximal benefit to the evolution of the design product. These goals can be accomplished by viewing measurement as a process involving the four phases shown in Figure 2.

Naturalist Phase

This phase involves understanding the domains and tasks of stakeholders from the perspective of individuals, the organization, and the environment. This understanding includes not only people’s activities, but also prevalent values and attitudes relative to productivity, technology, and change in general. Evaluative assessments of interest include identification of difficult and easy aspects of tasks, barriers to and potential avenues of improvement, and the relative leverage of the various stakeholders in the organization.

Marketing Phase

Once one understands the domain and tasks of current and potential stakeholders, one is in a position to conceptualize alternative products or systems to support these people. Product concepts can be used for initial marketing in the sense of determining how users react to the concepts. Stakeholders’ reactions are needed relative to validity, acceptability, and viability. In other words, one wants to determine whether or not people perceive a product concept as solving an important problem, solving it in an acceptable way, and solving it at a reasonable cost.

Engineering Phase

One now is in a position to begin trade-offs between desired conceptual functionality and techno- logical reality. As indicated in Figure 2, technology development will usually have been pursued prior to and in parallel with the naturalist and marketing phases. This will have at least partially ensured that the product concepts shown to stakeholders were not technologically or economically ridiculous. However, one now must be very specific about how desired functionality is to be provided, what performance is possible, and the time and dollars necessary to provide it.

Sales and Service Phase

As this phase begins, the product should have successfully been tested, verified, demonstrated, and evaluated. From a measurement point of view, the focus is now on validity, acceptability, and viability. It is also at this point that one ensures that implementation conditions are consistent with the as- sumptions underlying the design basis of the product or system.

The Role of Technology

It is important to note the role of technology in the human-centered design process. As depicted in Figure 2, technology is pursued in parallel with the four phases of the design process. In fact, technology feasibility, development, and refinement usually consume the lion’s share of the resources in a product or system design effort. However, technology should not drive the design process. Human-centered design objectives should drive the process and technology should support these objectives.

Organization for Measurement

Table 1 illustrates how the seven measurement issues should be organized, or sequenced, in the four phases. Framing an issue denotes the process of determining what an issue means within a particular context and defining the variables to be measured. Planning is concerned with devising a sequence of steps and schedule for making measurements. Refining involves using initial results to modify the plan, or perhaps even rethink issues and variables. Finally, completing is the process of making outcome measurements and interpreting results.

Table 1 provides a useful context in which to discuss typical measurement problems. There are two classes of problems of interest. The first class is planning too late, where, for example, failure to plan for assessing acceptability can preclude measurement prior to putting a product into use. The second class of problems is executing too early, where, for instance, demonstrations are executed prior to resolving test and verification issues, and potentially lead to negative initial impressions of a product or system.