CONCEPT SELECTION

4054 Lecture Notes

W. Durfee Sept-99

(w/ help from J. Davidson, T. Chase)

 

 

Bring:

 

SIT IN YOUR TEAMS

 

Ref: Ulrich and Eppinger, Chapt. 6

Web notes

 

BACKGROUND

"Concept Selection" is picking the idea(s) which best satisfy the Product Design Specification (PDS)

Stage in design process: After (1) understanding customer needs, (2) developing PDS, (3) generating many concepts. Before detailed design.

You are selecting among choices constantly in design process. If you donít have many choices to choose from at every stage of the design process, your process is bad.

 

Selection is an iterative process. (1) May need new or modified concepts, (2) May need more info to proceed.

Think of concept selection as weeding out bad ideas, not as trying to pick the "best" idea. The latter is too hard. Selection should be a narrowing process, not a competition.

Put yourself in customers shoes for the selection. No good if you pick based on your own perceptions since you arenít buying lots.

Wait a day or two after any branstorming sessions before running a selection process.

Stay flexible

You NEVER have enough information to do it properly...but often you canít wait.

 

 

- MUST follow a structured selection process. Ensures that:

* Customer needs drive process

* All criteria are considered

* Choices are documented

* Choices can be defended later in the design cycle

* Process immune from influence of champion

 

- The ways NOT to make selection decisions:

* "Gut feel"

* The boss says it has to be this way

* Let a single customer decide

* Team member who is strong champion of one idea

* Influence of experienced designer

 

 

 

PRELIMINARY STEPS

Agree on the criteria by which concepts will be evaluated. Get entire team (including sponsors) to buy in. Weight? Power? Time-to-Market? Cost? Comfort? Patentable?

Criteria should be CUSTOMER DRIVEN!

Present all concepts in same level of detail

- All drawn by the same hand

- All detailed to same level

If not, most "polished" concept will win

 

 

MULTI-VOTING (Red-Dot/Green-Dot) METHOD

Good, informal method to get team membersí gut feel.

Give everyone on team 5 red dots and 5 green dots (Avery sticky dots, 3/4 in. dia.)

List (or sketch) concepts on large paper using dark, fat markers. Post on wall.

Members stick their green dots on concepts they like. Can use more than one dot per concept. Use red dots for concepts which should be eliminated.

Not recommended for actual selection, but is a good way to see where people stand.

 

PUGH CONCEPT SELECTION (Preferred Method)

 

Matrix

Pick "datum" concept

Three level comparisons of criteria to datum: +, 0, -

Use to eliminate/combine ideas

Very powerful method with appropriate mix of qualitative/quantitative

See Ulrich and Eppinger pp 113-116 for details.

 

 

QUANTITATIVE SELECTION MATRICES (Useful, but problematic)

Matrix

Weight criteria

Score concepts against criteria

Calculate weighted total score for each criteria

Problem 1: numbers imply a degree of confidence which cannot be justified.

Problem 2: results very sensitive to who does the scoring.

See Ulrich and Eppinger pp 117-120 for details

 

REFLECT (Always)

Do the results make sense?

Does the team feel comfortable with the results?

Does the project sponsor feel comfortable with the results? (VERY important)

If you did it again, would it come out differently?

Do you need more information to do it properly?

Do you have to select now because of project deadlines even if you donít have full information?

 

 

 

Exercise SELECTING A WRITING INSTRUMENT

In your team.

Everyone (record on board): Customer needs for writing instruments:

* smooth writing, stays sharp, lasts long, can carry in my pocket, looks kool, comfortable, can erase, inexpensive, small.....

What are the criteria? (prune to 6)

Within your group, find N different writing instruments (pens, pencils if not enough pens), where N = group size.

Set up a Pugh chart and evaluate, one person per criteria.

Now set up "secret" quantitative charts, one per team member. Weight criteria, score and total. Compare "winners" among your team. Same/different?

Note: in this exercise, you had the products (concepts) to test. For design, you may or may not have more than just an idea or drawing. Makes it harder to do the assessment.

Turn in your quantitative sheet at end of class

 

 

ADDITIONAL NOTES

- If you are pressed for time, or want to make a first cut, go around the table and have each member vote on his or her favorite and explain why. All members must have a chance to speak up in this process. Be wary of making a final decision using this process, however, many times it is a way of doing an initial screening. Remember also that this may be the first time people have a chance to critique the ideas. Can do the same thing by voting on the concepts silently. Post all ideas on walls. Each team member "brands" their one or two top concepts. Keep the populuar ones, eliminate the weak ones. Both of these methods lack the group discussion that helps to make careful, appropriate decisions.

 

- Think of selection as "controlled convergence" (Pugh, 1981), that is you are doing a multi-step process to narrow down the choices. After each selection stage, combine and strengthen existing concepts and use the selection results to generate new, stronger concepts for the next selection round. Often, more info must be gathered before the next selection round is possible.

- It's an iterative process. New concepts can be added to the selection process at any time.

- When deciding among concepts, be sure that all concepts are presented at approximately the same level of detail. For example, it is unfair to compare a crudely sketched concept fragment with a detailed working drawing. Equality should be in scope as well as presentation (it's the same phenomena as a word-processed document generally commanding more respect than a hand-written memo). Representing each concept as a thumbnail sketch goes a long ways towards leveling the playing field here. If level of detail matters in concept selection, express it as a selection criteria rather than skewing the presentation.

 

- Structured selection is generally best done in a matrix format. Selection criteria listed down left column, concepts listed along top row. Depending on complexity, may be up to 100 criteria and 20 concepts. If the matrix is overwhelming, generally it means the problem was not well divided into subtopics, or that decisions at the systems level are being driven by the detail level.

- Before proceeding, team must determine the selection criteria. These are the specific evaluation measures that the concepts will be judged on. Criteria should be CUSTOMER DRIVEN!!! Criteria should be developed by consensus of the team. Approval of selection criteria by all of the stakeholders is important. Stakeholders may include design team (engineering, marketing, sales, manufacturing), management (immediate and senior), customers. Detail of selection criteria depends on detail of concept. For selecting system level concepts, criteria might be quite abstract.

- Choose a medium for recording the matrix. Whiteboard, large rolls of paper, electronic chaulkboard, projected overhead foil (or computer screen) are possibilities. Should be one large-format medium rather than separate sheets of paper or separate video screens. Want everyones eyes on a common point making common decisions rather than multiple, private decisions.

- Matrix selection methods can be unweighted ("Pugh concept selection") or weighted. Which you use depends on level of abstraction of concepts and criteria. Unweighted is generally better. Weighted implies a degree of quantitative confidence that generally cannot be defended.

- Pugh concept selection: <*** refer to lecture ***>

- Weighted selection <*** refer to lecture***>

 

- After selection process, the team should sit back and take a fresh look. Does the selection match intuition? If not, why not? Was there anything that was missed? Does the team feel comfortable about the choices? Remember, these are important decisions that are being made. You're putting the life of the project into your chosen concept. It better be the best or else you will be in trouble down the line.

 

- Often, it is during the selection process that a closer evaluation of the PDS occurs. Often you are in the situation where all of the concepts look lousy. This can be the result of two problems: (1) poor concept selection process, (2) PDS is inappropriate. For example, the team may have to decide to go with a reduced version of the product to get it out now, because meeting all of the specs requires new technologies to be developed (i.e., the spec cannot be reached).

 

EXAMPLES OF WEIGHTED AND UNWEIGHTED SELECTION CHARTS

 

Weighted

Wght

Battery

Pneumatic

Corded Electric

Int. Combustion

Ease of Use

40

Setup

20

5

1

2

4

Operating

20

4

2

2

3

Weight

30

2

4

3

2

Manufacturing Cost

20

3

4

3

2

Time to Market

20

3

4

3

3

Total Score

360

340

290

300

Rank

1

2

4

3

 

Unweighted

Battery

Pneumatic

Corded Electric

Int. Combustion

Ease of Use

Setup

+

S

+

+

Operating

+

S

S

+

Weight

-

S

-

-

Manufacturing Cost

-

S

-

-

Time to Market

-

S

-

-

# of Pluses

2

0

1

2

# of Minuses

3

0

3

3

Keep?

Y

N

N

Y