Georges M. Fadel
Clemson University
August 1997

About a month ago, I had the opportunity to attend a conference on OR/Management Science which included such topics as OR and group decision making, decision making in groupware environments, OR/MS in the finance and the planning of financial services, Decision Support Systems, graphical models in decision analysis, multi-criteria decision analysis, etc.. I listened to many papers on decision software, utility theory, weighted criteria, etc.. What I gathered from the talks was that a wealth of information and research exists on decision making for many fields, and I wondered why we were looking at decision based design without much input from the mathematicians, financial gurus, management people and the such. (Maybe some of us are??)

One interesting observation, (maybe biased by the talks selected) was that even though many techniques exist to deal with multiple criteria, the linear weighted method is overwhelmingly the brute force used to come up with a single objective. So are our applications in design so different from other fields that many want to be able to combine objectives in a non-linear manner?

Another interesting topic brought up the topic of timing of decision making. Several examples were given: for instance, in certain cases, a decision made early on narrows the choices subsequently, and eases the solution process. In other cases, delaying the decision may allow the decision maker to make better decisions. In yet other cases, the timing may not play any role. The speaker gave the example of environmental problems or nuclear accidents and the need to make decisions for evacuation. By waiting, the situation may be better assessed and the decision delay might prove "cost effective".

So looking at the design process, I ask myself, where do we need to make decisions? When should we make the decisions? Can we assess the effect of some decision on the overall design? Is decision making in design tightly related at least at some stage to the definition of the way to consider multiple criteria, combining them, or selecting one at the expense of others, or ...

I would like to consider a design problem, use some methodology to attack the problem, and identify the different stages of the design process where decisions have to be taken. The next step would be to try assess the impact of decision making. First, what percent of time (in the complete design process) do we spend making decisions? Are the decisions considered at various stages of the process similar? Is the effect of decision a function of when that decision was taken, of what it applied to (Generation of objective(s), constraints, selection of alternatives, managerial issues, financial issues, marketing issues, ...), of the degree of uncertainty, of risk,.. Can we classify decision making in design? Can we from these questions determine the type of decisions and their timing that will lead to better designs?

Finally, if, or better, when we learn more about the decision process in design, I would like to how will this help us teach design and make better designers of our students.