Concept Generation - University of Iowauser.engineering.uiowa.edu/~bme083/lecture/lecture13_030503.pdflMany methods for concept generation are available, but no single method is best.

Concept Generation

l Concepts are the means for providing function.– Any form that gives an indication how the function can be achieved.– What to do → Function vs. How to do → Concepts (forms)– Remember that the idea is often not original.

l Many methods for concept generation are available, but no single method is best.

– A good designer is familiar with these methods and uses them, or a combination of them, as needed.

Basic Methods for Concept Generation

Morphological Methodl This technique uses the functions identified

to foster ideas.– Powerful method that can be used formally

or informally as part of everyday thinking– 2 step approach

l Step 1 - Developing Concepts for Each Function:

– Goal is to find as many concepts as possible that can provide each function identified in the decomposition.

• How to store mechanical energy: springs, elastomers (rubbers or plastics), etc.

• If there is a function with only one conceptual idea, this function must be re -examined.

– Situations explaining the lack of more concepts.

• The designer has made a fundamental assumption.

• The function is directed at how, not what. -e.g. “store energy in coil spring” rather than “store energy”

• Domain knowledge is limited.– Keep the concepts as abstract as possible and

at the same level of abstraction for better comparison of developed concepts.

• The force required for moving an object can be provided by a hydraulic piston, a linear electric motor, the impact of another object, or magnetic repulsion.

– Refined mechanical components vs. basic physical principles.

Morphology for BikeE Suspension System

Morphological Methodl Step 2 - Combining Concepts:

– to combine these individual concepts into overall concepts to meet all the functional requirements.

• Select one concept for each function and combine those into a single design.

– Pitfalls:• This method may generate too many

ideas.• It erroneously assumes that each function

of the design is independent and that each function satisfies only one function.

• The results may not make any sense.

l Concept generation process is the time that sketches and begin useful.

– We remember functions by their forms– Only way to design an object with any

complexity is to use sketches to extend the short-term memory.

– Sketches made in the design notebook provide a clear record of the development of the concept and the product.

Combined Concepts for BikeE Suspension System

Logical Methods for Concept Generationl The Theory of Inventive Machines, TRIZ:

– Developed by Genrikh Altshuller (a ME engineer, inventor, and Soviet patent investigator) in Soviet Union in the 1950s based on patterns found in patented ideas

– Goal of TRIZ:• Find the major contradiction that is making the problem hard to solve, then• Use TRIZ’s 40 inventive ideas for overcoming the contraindication

– With TRIZ, we can systematically innovate; we don’t have to wait for an inspiration or use the trial and error common to other methods.

l Axiomatic Design:– Evolved in MIT by Prof. Nam Suh in an effort to make the design

process logical.– 1st Axiom:

• Maintain the independence.– Then, a change in a specific design parameter should have an effect only on a single function.

– 2nd Axiom:• Minimize the information content of the design.

– The simplest design has the highest probability of success and is the best alternative.

Concept Evaluationl How to choose the best of the concepts generated for

development into a quality product?– Goal is to expend the least amount of resources on deciding which

concepts have the highest potential for becoming a quality product.– It is difficult to evaluate concepts, or to choose which concepts to spend

time, particularly when we still have very limited knowledge and data on which to base this selection.

• Design is learning, and resources are limited.

l Techniques for systematic evaluation of rough concepts.– Evaluation implies both “comparison” and “decision making.”

• It is the comparisons between alternative concepts and the requirements that they must meet that gives the information necessary to make decisions.

– For all design decisions:• Itemizing the alternatives and the criteria for their evaluation• Comparing the alternatives to the criteria to each other

– For comparisons:• Alternatives and criteria must be in the same language and they must exist

at the same level of abstraction

Concept Evaluation TechniquesBe ready during concept evaluation to abandon your favorite idea, if you can not defend it in a rational way.

Abandon if necessary “the way things have always have been done around here.

Information Presentation in Concept Evaluation

Design Evaluation Cycles

Evaluation based on Feasibility Judgmentl Three Immediate Reactions of a Designer as a concept is generated based on

designer’s “gut feel”:– It is not Feasible.– It might work if something else happens.– It is worth considering.

l Implications of Each of these Reactions:– It Is Not Feasible.

• Before discarding an idea, ask “Why is it not feasible?”• Make sure not to discard an idea because:

– a concept is similar to ones that are already established, or– a concept is not invented here (less ego-satisfying).

– It is Conditional.• To judge a concept workable if something else happens.• Factors are the readiness of technology, the possibility of obtaining currently

unavailable information, or the development of some other part of the product.

– It is Worth Considering.• The hardest concept to evaluate is one that is not obviously a good idea or a bad one,

but looks worth considering.• Such a concept requires engineering knowledge and experience. If sufficient

knowledge is not immediately available, it must be developed using models or prototypes that are easily evaluated.

Evaluation based on GO/NO-GO Screening

l Measures for deciding to go or no-go:– Criteria defined by the customer requirements:

• Absolute evaluation by comparing each alternative concept with the customer requirements.

• A concept with a few no-go responses may be worth modifying rather than eliminating

• This type of evaluation not only weeds out designs that should not be considered further, but also helps generates new ideas.

– Readiness of the technologies used:• This technique refines the evaluation by forcing an absolute comparison

with state-of-the-art capabilities.• 6 Measures for a Technology’s Maturity:

– Are the critical parameters that control the function identified? – Are the safe operating latitude and sensitivity of the parameters known?– Have the failure modes been identified?– Can the technology be manufactured with known process?– Does hardware exist that demonstrates positive answers to the preceding four questions?– Is the technology controllable throught the product’s life cycle?

• If these questions are not answered in the positive, a consultant or vendor is added to the team.

Evaluation based on a Basic Decision Matrix

l Decision-Matrix Method (or Pugh’s Method):

– effective comparison of alternative concepts (basic form Table 8.2)

– Iteratively test the completeness and understanding of requirements, rapidly identifies the strongest

l Step 1: Choose the criteria for Comparison.– Criteria are the functional requirements and

engineering specification determined in QFD.– The concepts must be refined enough to

compare with the engineering targets for evaluation (mismatch in the level of abstraction).

l Step 2: Develop Relative Importance Weightings.

– Step 3 in QFD should provide the data for relative importance.

l Step 3: Select the Alternatives to be Compared.

l Step 4: Evaluate Alternatives.– Relative evaluation among alternatives

l Step 5: Compute the Satisfaction.

**This method is most effective if each member performs it independently and the individual results are then compared.

- have BDM in individual notebook.

Decision Matrix for Energy Management System

S indicates “Same as datum”

Robust Decision Making

l Robust decision refers to make decisions that are as insensitiveas possible to the uncertainty, incompleteness, and evolution ofthe information that they are based on.

l For robust decision making, we need to improve the method used to evaluate the alternatives (step 4 in decision-matrix method).

l Word Equations used for Robust Decision Making– Satisfaction = belief that an alternative meets the criteria– Belief = knowledge + confidence

• Belief is the confidence placed on an alternative’s ability to meet a target set by a criterion, requirement, or specification, based on current knowledge.

• Belief (virtual sum of knowledge and confidence) can be expressed on a “belief map.”

Belief Map

Knowledge = a measure of the information held by a decision maker abouta feature of an alternative defined by a criterion

Mea

sure

of C

onfid

ence

:Belief = p(k) x p(c) + (1 – p(k)) x (1 – p(c))

Belief Map showing the ProbabilitiesBelief Model:

Belief = p(k) x p(c) + (1 – p(k)) x (1 – p(c))

Extreme Cases of Belief Map

p(k) = 1.0p(c) = 1.0

Certainly work

p(k) = 1.0p(c) = 0.0

Certainly not work

p(k) = 0.5p(c) = 0.0

Pessimist corner

No more info. For evaluation

p(k) = 0.5p(c) = 1.0

Optimist corner

Not clear

Belief Map Example

Point A: p(k) = 0.8 (informed knowledge) and p(c) = 0.65 (high confidence); belief = 0.59He/she has a belief of 59% that the concept will meet the functional requirements.

Point B: p(k) = 0.9, p(c)=0.25, Belief = 0.30

Arrow shows the evolution (With increasing knowledge, confidence and belief decrease.).

Belief Map to Choose Where to Eat

Criteria:delighted (p(c) = 1.0) if < $5.00disgusted (p(c)= 0.0) if > $10.00fast food = $6.00 (sure)Family diner = $9 (most likely)Chinese = $7 (little knowledge)

Fast food should be chosen for the best probability for satisfaction.

* Choice can be altered with inclusion of other criteria, such as food quality, taste, nutrition, etc.

Advanced Decision Matrixl Steps 1 through 3: same as the Decision Matrix Methodl Step 4: Evaluate Alternatives

– Use a belief map for comparison– If little is known or the evaluation result is that the alternative possibly meets the

criterion, then belief = 0.5

l Step 5: Compute Satisfaction– Satisfaction = Σ (belief x importance weighting)

• Max satisfaction = 100 (evaluator is 100% satisfied.)

Product Safety and Liability

l Safety is best thought of early in the design process.– Product safety and liability are often overlooked until late in the project.

l Product Safety:– Design for safety means ensuring that the product will not injury or loss.– 2 Safety Issues:

• Who or what is protected from injury or loss?• How is the protection actually implemented in the product?

– 3 Ways to Institute the Safety:• Design safety directly into the product

– device poses no inherent danger during normal operation

• Add protective devices to the product (automatic cut-off switch).• Warning of the dangers inherent in the use of the product.

l Products Liability:– Special branch of law dealing with alleged personal injury or property

or environmental damage resulting from a defect in a product.

Products Liabilityl 3 Different Charges of Negligence against Designers:

– The product was defectively designed.• Failure to use state-of-the-art design considerations.• Improper analyses, use of poor materials

– The design did not include proper safety devices.– The designer did not foresee possible alternative uses of the product.

l What to do for protecting ourselves as a designer?– Keep good records to show all that was considered during the design

process.– Use commonly accepted standards when available.– Use state-of-the-art evaluation techniques for proving the quality of the

design before it goes into production.– Follow a rational design process so that the reasoning behind design

decisions can be defended.

Hazard AssessmentWhen a potential hazard is identified,1. Estimate the frequency of occurrence

(Table 8.6);2. Estimate the consequence of occurrence

(Table 8.7);3. Predict the hazard index using the hazard-

assessment matrix (Table 8.8);

If the hazard index is low (undesired or unacceptable), further consideration must be given on the design.