Understanding – really understanding – a design problem is half the battle in finding a good design solution.
Problem analysis is a set of analytic tasks meant to increase the designers' understanding of an unbalanced situation, for the sake of designing a change to the situation that will have better balance.
In other words, problem analysis involves developing a set of requirements that will be satisfied by any suitable design intervention, and only by suitable design interventions.
With reference to Simon's definition of designing, the requirements will quantify what we mean by a “preferred situation” that must be achieved.
Situations are problematic only for people - nature doesn't have any “problems.” To understand a situation, you have to understand the people and other agents who find the situation dispreferred.
It is vitally important that your understanding of the problem be properly documented, because:
If you cannot properly analyze a design problem, then you are quite likely to come up with a “bad” design - where a “bad” design is one that doesn't address the imbalances in a situation. Since we learn best from our mistakes, it is quite useful to look at bad designs and think about what's wrong with them and how they might be fixed.
Some examples of bad designs (or badly analyzed design problems) include:
Finally, there is one more reason for performing a detailed and well-documented problem analysis: it helps “synchronize” the thinking of all the team members. At the very outset of a design project, the problem is almost invariably defined in quite vague terms. Everyone on the team has a different experience base, different attitudes and perspectives, different amounts of knowledge, etc. In the situation of a vague problem, this means, different team members will very likely end up with different mental models of the problem. This is a problem, because it means each team member - if left to their own devices - will address a slightly different problem than every other team member. This will end badly for everyone.
By conducting problem analysis in a team setting, all team members will (if it's done right) end up agreeing quite precisely on the actual imbalance(s) that have to be addressed. That is, all team members will have the same mental model of the problem - and so the resulting design intervention is that much more likely to address the imbalances well.
Assuming you have completed the project initialization phase, there are two main steps to problem analysis:
Developing a set of requirements will probably induce questions (a) that you had not yet thought of, yet (b) the answers to which will improve the requirements. This results in an iterative process, composed of alternating phases of exploration, during which more information is gathered or identified, and analysis, during which information is pared down, made more concise and precise, and formalized.
The basic steps of the process (per the Design Roadmap) are as follows.
The formal deliverable of the problem analysis stage is a product requirements specification.
Many imbalances do not relate directly to engineering concerns. Since our responsibilities and obligations are only to the engineering aspects of a design problems, we need to be able to pick out those technical aspects and focus on them. However, the technical aspects of a specific situation are coupled to other non-technical aspects in the broader societal situation. This is why we cannot ignore completely the non-technical aspects of situations.
Indeed, our proposed resolutions for the technical aspects of unbalanced situations are constrained by the non-technical issues, because we, as engineers, have no control over the non-technical issues - that is, in the same way that we as engineers are supposed to have control over the technical issues, we must accept that there are other kinds of experts who have control over the non-technical aspects.
Knowing where to draw the boundaries between the aspects on which you need to focus and all the other aspects can be quite vague. There may well be some aspects about which your team cannot decide. In such cases, do not hesitate to seek guidance from your instructors.
Also, even with all the best intentions and meticulous execution of this stage, it is not guaranteed that everyone will agree. The best one can expect, usually, is “informed compromise” [AEF00] about the problem that has to be solved. Managing those compromises will become a key element of the process as the design proceeds.