1.6 Designing as heuristic problem-solving

Generally, solving design problems is different from solving puzzles or mathematical equations. One major difference is that design problems are often not well-specified. This is discussed in more detail in the next section, but one of the primary reasons is the complexity of factors which you began to explore in SAQ 2. It means that the properties of the object that the designer is supposed to produce are often not very clear; if they were entirely specified, the designer's job would almost have been done!

Another difference is that design problems don't usually have a single 'correct' solution. Generally there are many possibilities. Sometimes if the requirements are over-specified there may be no solutions. For example, 'Design an aeroplane for 150 passengers that will cost less than £100,' or 'Design a car which will give complete protection to its passengers in any high-speed accident.'

Consider an under-specified problem, such as 'Design me a house for two adults and three children costing less than five times my salary.' There could be millions of possible designs meeting this specification; and there would be millions more designs not meeting the specification. How can the designer find a good design from all these millions of possibilities? One approach would be to try looking at them all, and judging which is the best. This is, of course, totally impractical. In general there is no formula which leads to a good solution, and designers have evolved heuristics for solving their problems. A heuristic is a rule or procedure which works most of the time, but sometimes fails. For example, one of the best ways to predict the weather tomorrow is to say it will be same as today. Of course this weather prediction heuristic fails sometimes: you may judge for yourself whether or not it has a pretty good record.

One heuristic used in solving problems with many possible solutions is to reduce the number of options. So the architect designing my house might ask me a few questions, such as whether I would like it to be made of brick, or how many bedrooms I would like, and so on. Each answer would reduce the set of possibilities dramatically. By asking the right questions, the architect could rapidly weed out the majority of 'non-solutions', and begin to investigate some of the remaining possibilities.

Another heuristic used by designers is to look at previous designs to see if there is already a solution to the problem. Often they find solutions to similar problems, which can be adapted to the design problem in hand.