General Reminder of Problem Solving Approach
Engineer: one who sees opportunity where others see problems
Engineers are problem solvers who use math, physics and other sciences combined with experimental work, and who must take into consideration a wide range of specifications, including the requirement of cost, to solve problems and provide solutions and products for the marketplace.
The engineer's problem solving approach a taught at Thayer School:
- Define the Problem: If given a problem, re-define it in your own terms -- beware! Problem statements provided by clients will:
- have client's bias in the statement
- typically suggest an approach and/or solution which is sub-optimal (implied solution)
- lack a complete understanding of the state of the art.
- Generate Specifications: Initially these may be quite general and focused on the problem and not specific to a solution. Typical specs are: size, weight, speed, cost, and accuracy; BUT, specs must also include: environmental and social impacts, reliability, maintainability, producibility, safety, noise, etc.
NOTE: Once a specific solution has been chosen for analysis, specification specific to that solution are necessary.
- Generate Alternatives: The use of brainstorming* techniques or Synectics is suggested here as a mechanism for producing large numbers of clever, creative solution paths which are non-intuitive and therefore potentially novel. Expect to make several approaches to the problem before you find one that works. For large or intractable problems try segmentation techniques to break the large problem into smaller, easier ones to handle.
*Brainstorming requirements: It must be fun or there will be no creativity. It should be unstructured, loud, collaborative, funny, outrageous and uncontrolled. There must be NO criticism of any idea; nothing is too crazy. There should be no analysis of any kind at this time. One group member should act as secretary to take down all ideas. Sometimes analogies will help.
- Analysis of Alternatives: You will have generated a large number of potential solutions to the problem (alternatives) so try to group them by type. Within a type, look for the critical factor which will determine whether or not it has potential. More research into the state of the art is generally useful at this stage. A first-cut analysis is also worthwhile and should be the fast, back of the envelope type -- possibly a simple experiment is quicker for some ideas.
- Selection of Alternatives: You should choose one to five of the most promising alternatives from the analysis just carried out and research them thoroughly using all of the resources of the library as well as personal and telephone contacts. As you learn more about the potential for your alternative, make a list of the specifications required for success. These specifications are a subset of those developed in Step 2 and are specific to the alternative.
- Iterate Steps 2 through 5: Always keep in mind "the problem." Make yourself a good paper trail so that if an alternative doesn't pan out, you will know exactly what to do next.
Learn all you can about the problem. Use literature, patent and market searching techniques and the telephone; you might be surprised how much free insight you can obtain.