DAMADV - Design

The Design Phase of the DMADV model is all about creating a detailed design for the high-level design developed in the Analyze Phase. It involves constructing the design; creating operational definitions for each detailed CTP; validating the measurement systems for detailed CTP; estimating the capability for each detailed CTP; and preparing a control and verification plan. At the end of the Design Phase, there is no doubt about exactly what is being developed; it is a welldefined concept.

• Use the Design Phase of the DMADV model to move from the highlevel design developed in the Analyze Phase to a detailed design.
• Use various tools and methods to predict the sigma level of the detailed design.
• After the Design Phase, build and pilot test a prototype of the product, service, or process under study.
  • Additionally, develop plans for full-scale production of the product or delivery of the service.

The Design Phase has five steps: 1 Constructing a detailed design (converting CTQs and high-level CTPs into detailed CTPs). * CTPs can exist at several levels of detail. 1 Operationally defining the detailed CTPs. 1 Validating the measurement systems for each detailed CTP. 1 Establishing the baseline capabilities of the detailed CTPs. 1 Preparing a verification plan.

The inputs from the Analyze Phase to the Design Phase of the DMADV model are the following (Git06, ch7.2)

• A high-level design (including CTQs and high-level CTPs for
  • product
  • service
  • processes
  • facilities
  • equipment
  • materials
  • supplies
  • information
  • people
• A set of design scorecards for the best high-level design.

Steps of the design phase

(Git06, ch7.1, p209)

constructing a detailed design

$ Developing the elements of the detailed design: $ Determining the details for the best design considering alternative detailed designs, selecting the best alternative detailed design, developing the specifics of the best detailed design, procuring the necessary materials, etc: $ Identifying targets for the detailed CTPs by calculating nominal values and specification limits and identifying control points and measurement systems.:

Creating operational definitions for detailed CTPs

validating the measurement systems that will be needed to audit performance of the detailed CTPs.

estimating the capability of the selected detailed design to meet the targets established for each CTP.

$ capability analysis: $ FMEA: $ Simulation: $ Design scorecard: $ Design review:

preparing a control and verification plan

• considers control charting or performance audits of the detailed CTPs
• identifies process owners
• identifies stakeholders
• documents the detailed design

Constructing a detailed design

Developing Elements of the Detailed Design; move from a high-level design to a detailed design by moving from general CTQs to high-level CTPs to detailed CTPs (e.g., dimensions, brand names, etc.).

use QFD to translate CTQs and high-level CTPs into detailed CTPs

QFD Methodology Flow

Quality Function Deployment for services has two basic matrices: $ Matrix 1: "Voice of the Customer" by "CTQs and CTPs" matrix: The "Voice of the Customer" by "CTQ" matrix is constructed using the steps discussed in detail in Section 5.7. $ Matrix 2: "CTQs and CTPs" by "Detailed CTPs: The CTQs defined in Matrix 1 become the "rows" that are listed down the left side of Matrix 2 along with priorities (based on the importance ratings from Matrix 1) and target values. Matrix 2 is prepared in the same manner as Matrix 1.

| CTQs and High-Level CTPs | Detailed CTPs ||||| Importance Ranking | | | Oven | Refrigerator |Large desk | Desk char | ... | ^ | | Kichenette | 9 | 9 | 0 | 0 | ... | | | Office Area | 0 | 0 | 9 | 9 | ... | | | ... | ... | ... | ... | ... | ... | | | Un-normalized weights | | | | | | | | Normalized weights | | | | | | | | Specific detailed CTPs | Whirlpool model 123/A | Frigidaire Model C54-X | | | | |

Remember, QFD requires multiple functional disciplines to adequately address all of its components.

Develop detailed CTPs for CTQs and High-Level CTPs

Language of CTQs and CTPs

(this seems to define/describe the USL etc.)

$ Specifications state a boundary, or boundaries: that apply to individual units of a product or service. $ An individual unit of product or service is considered to conform: to a specification if it is on or inside the boundary or boundaries; this is the goal post view of quality. $ Individual unit specifications are made up of two parts: together they form a third part $ nominal value: This is the desired value for process performance mandated by the customer's needs. Ideally, if all quality characteristics were at nominal, products and services would perform as expected over their life cycle. $ tolerance. A tolerance is an allowable departure from a nominal value established by design engineers that is deemed nonharmful to the functioning of the product or service over its life cycle.

Flow-UP CTPs

Dimensions are created by the assembly of the steps in a service or component parts of a product.

These created (flow-up) dimensions have statistical distributions. For example, if nine services (CTPs) are put together to form a one-stop-shopping service (CTQ), the distribution of the one-stop-shopping service (CTQ) is a newly created dimension.

If component steps or parts are assembled so that the individual step or component dimensions are added to one another, the average dimension of the assembly will equal the sum of the individual component average dimensions, assuming the assembly adds no measurable dimension.

The preceding law holds only if the processes generating the steps or parts are independent of each other (i.e., the results of one do not impact the results of the other) and in statistical control (i.e., they exhibit a predictable system of variation). Independence among components is not always the case. For example, suppose the time it takes to have a meal served at a restaurant is composed of the three steps: Step 1: take the order; Step 2: cook the order; and Step 3: deliver the order. If the first step is done incorrectly, there will be a substantial impact on the second and third steps.

$ Law of the Addition of Component Dimension Standard Deviations: Spp= SQ(Sp^2+Se^2+Sc^2) * the standard deviation of the assembly will be the square root of the sum of the component variances, regardless of whether the components are added or subtracted from each other. * This law applies to assemblies in which the component steps and parts combine linearly and are statistically independent.

Flow-Down CTQs

Flow-down is the design function in which a CTQ is broken down into high-level CTPs, or high-level CTPs are broken down into detailed CTPs.

For example, a "design process" consists of 3 steps: Step 1 is collecting "Voice of the Stakeholder" data for a design, Step 2 is preparing the design, and Step 3 is delivering the design to the client. Ideally, the client wants the design in exactly three weeks (Nominal = 3 weeks), but no later than four weeks (USL = 4 weeks).

deviation needed is defined as:

• 4 weeks = 3 weeks + 6* SST where SST = short-term standard deviation of the design delivery process (zero sigma shift in the process mean). In other words,
• SST = 1/6 week

Assume a short term normal distribution with a mean of M and a standard deviation of SST. The Voice of the Customer has stated an USL = M + 6SST.

Because overall delivery time for the design = step 1 + step 2 + step 3, the transfer functions for the flow-down mean and standard deviation are the following:

Consequences of Failure to Flow-Down

Three options exist if the detailed designs cannot deliver the detailed CTPs required by the CTQs and high-level CTPs, given the available budget. $ needs to be redesigned to achieve design and/or budgetary targets: A helpful technique for checking to determine if the designed or redesigned detailed CTPs meet the specifications for all CTQs and CTPs is to perform a flow-up capability analysis. * Flow-up capability analysis creates the opportunity for designers to determine if flow-down targets and tolerances actually achieve CTQs and higher-level CTPs targets and tolerances. * The statistical stuff. $ design trade-offs must be made among the detailed CTPs to achieve the design and/or budgetary targets for all CTQs and high-level CTPs: Design tradeoffs answer the question: Can flow-down targets and tolerances achieve higher-level targets and tolerances? $ The third consequence is that the design is abandoned in favor of other, more promising, uses for the resources that have been earmarked for the design in question:

Create a comprehensive set of detailed CTPs

$ Product or Service Detailed CTPs: These detailed CTPs include all detail necessary to manufacture products or deliver services; for example, descriptions of services with targets and tolerances that consider all stakeholder voices $ Process Detailed CTPs: These detailed CTPs include all high-level and detailed flowcharts and/or process maps necessary to manufacture products, deliver services, or provide process outputs. $ Human System Detailed CTPs: These detailed CTPs include job descriptions, skill requirements, educational history, experience, hiring policies, training policies, supervisory processes, performance enhancement processes (including extrinsic motivators), and management style $ Information System Detailed CTPs: These detailed CTPs include all hardware, software, and humanware required to manufacture products, deliver services, or provide process outputs. $ Equipment and Tools Detailed CTPs: These detailed CTPs include descriptions, instructions and drawings, set-up procedures and specifications, capabilities, and operating conditions. $ Material and Supply Detailed CTPs: These detailed CTPs include descriptions, instructions for use, specifications, conditions for use, accounting information (e.g., billing policies), inventory policies, and replenishment processes $ Buildings and Grounds Detailed CTPs: These detailed CTPs include floor plans, and space layout drawings and models

Operationally define each detailed CTP

an operational definition promotes understanding between people by putting communicable meaning into words. An operational definition contains three parts(Git06,ch7.6, p224):

• a criterion to be applied to an object or group,
• a test of the object or group,
• a decision as to whether the object or group met the criterion
  • each CTQ and CTP must have a valid measurement system. A valid measurement system is a prerequisite for valid decision making.

Establish baseline capabilities for each CTQ and CTP

Background

• Baseline data needs to be collected to determine the stability and capability of each detailed CTP.
• The collection of baseline data requires a data collection plan.
  • The elements of a data collection plan include
  • a data collection repository or form,
  • a sampling plan (sample size, sample frequency),
  • sampling instructions (who, where, when, and how).

analyze the baseline data to answer the following questions: $ Does the baseline data for the detailed CTP exhibit any patterns over time: A run chart is used to study raw baseline data over time. $ Is the baseline data for the detailed CTP stable: Does it exhibit any special causes of variation? Control charts are used to determine the stability of a process. $ If the detailed CTP is not stable: (i.e., it exhibits special causes of variation), determine the location of the special causes of variation to take appropriate corrective actions to stabilize the detailed CTP. Again, you use a control chart to identify where and when special causes of variation occur. * However, you do not use a control chart to identify the causes of special variations. You use tools such as log sheets, brainstorming, and cause and effect diagrams to identify the causes of special variation. $ If the baseline data for the detailed CTP is stable, then you need to determine the characteristics of its distribution: In other words, what is its spread (variation), shape (distribution), and center (mean, median and mode)?

Estimate Process Capability for Appropriate Detailed CTPs

Process capability compares the

• output of a process ("Voice of the Process")
• customer's specification limits for the outputs ("Voice of the Customer"). A process must be stable (as shown in a control chart) to determine its capability.

There are two types of process capability studies: $ Attribute Process Capability Studies: Attribute process capability studies determine a detailed CTPs capability in terms of fraction defective output or counts of defects for a unit of output. The major tools used in attribute process capability studies are attribute control charts, and the tools discussed * Attribute process capability studies typically require an order of magnitude (i.e., 10 or more times) more data than measurement process capability studies, to gain similar confidence from the analysis. In general, an attribute process capability study should contain 20 to 25 samples and each sample should have between 50 and 100 units. Measurement analysis often uses samples of 4 to 6, which is 10% of the size indicated here for attribute data capability studies. This rule is based on experience, as well as statistical theory. $ Variables process capability: studies determine a process's ability to meet specifications for detailed CTPs. The major tools used in variables process capability studies are variables control charts and the tools discussed in Chapter 12. Variables control charts are used to stabilize a process so you can determine meaningful upper and lower natural limits. * Natural limits are computed for stable processes by adding and subtracting three times the standard deviation of the process to the process centerline. * As a general rule, the time period under study should contain at least 50 samples of between three and six units each(Git06, p234).

Natural limits ("Voice of the Process") and specification limits ("Voice of the Customer") are comparable quantities for stable processes because they are both measured with respect to the individual units of output generated by the process under study. There are four basic relationships between natural limits and specification limits for stable processes. Each relationship is portrayed using a stable normal distribution.

• The process's natural limits are inside the specification limits and the process is centered on nominal.
• The process's natural limits are inside the specification limits and the process is not centered on nominal.
• The process's natural limits are outside the specification limits and the process is centered on nominal.
• The process's natural limits are outside the specification limits and the process is not centered on nominal.

Capacity Analysis

Capacity analysis is a technique used to study the resource requirements and statistics for each detailed CTP in selected configurations of the detailed design.

• for example, waiting line or queuing statistics and bottleneck statistics.

1 develop and verify a discrete event simulation model of the detailed design to obtain an acceptable replica of the detailed design (detailed CTPs, CTPs, and CTQs). 1 design, conduct, and analyze statistical experiments using the simulation model to identify the settings of the detailed CTPs that optimize the CTPs and CTQs, as well as to estimate the resource requirements for the optimal detailed design.

Perform a FMEA of the detailed CTPs

| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | |Critical Parameter|Potential Failure Mode|Potential Failure Effect|Severity|Potential Causes|Occurrence|Current Controls|Detection|Before RPN|Recommend Action|Responsibility and Target Date|Action Taken|Severity|Occurrence|Detection|After RPN| |Accessible Roof|Fall or Jump|Death|5|Accident|1|None|5|25|Guard Rails|Architect|Guard rails in design|5|1|1|5|

Constructing detailed design scorecards

create scorecards for the detailed CTPs flowed-down from the CTP and CTQ scorecards in the Measure and Analyze phases.

• Part A presents the Voice of the Customer for each detailed CTP. The detailed CTPs listed in Part A are frequently an output of the second QFD matrix discussed on page 116. The Voice of the Customer for each detailed CTP includes the nominal value, the upper and/or lower specification limit(s), and a flow-down process sigma target.
• Part B of the design scorecard presents the simulated statistics for each detailed CTP; that is, the mean, standard deviation, DPU (using the USL and LSL), and the predicted process sigma for the detailed design.

Performing accounting analysis

Just as there are targets and specifications for product, service, and process characteristics in an overall design, there are budgetary requirements (constraints) for the product, service, or process characteristics. If the cost of developing and delivering a new design exceeds the benefits expected to be gained from implementing the new design, there are four decisions that can be made: 1 Do not implement the new design. * not a good design if the reason you are trying to implement a new design is to overcome issues or problems with your current design. Nonetheless, if it is clear that the cost to implement the design is far too high for the benefits achieved, it is legitimate for leadership to elect to forego the new design. 1 Investigate potential tradeoffs in the performance levels of the new design characteristics to decrease costs. * new design in phases (MGPP). a. The first few phases identify the major tradeoffs required to reduce the cost of the design. a. The next phases introduce new features or entirely new products, services, and processes that reduce the cost of the new design. * You can amortize the cost of implementing a phased design over a longer period of time; that is, the time frame represented by the phases 1 Investigate other alternative designs, or become more creative with respect to the characteristics of the current design. * alternative designs identified, but not used, from the Analyze Phase to find less costly designs, or less costly components for a new design. * If the overall costs of a new design are high due to so many design elements, perhaps an alternative design can be identified using a Pugh Matrix with fewer design elements. 1 Decide to go ahead with the new design for strategic reasons. * uses the corporate strategy to capture the market—for example, a new design becomes a loss leader for other designs.

The accounting and financial costs of a new design include:

• operating costs
  • Operating costs include the cost of labor, supplies, maintenance, and support for delivery of the new design.
• capital costs
  • associated with the investments so as to enable the design are often considered separately from operating costs. Cash flow is also a consideration. The difference between the revenue and gains from implementing the new design and the operating costs and the depreciation from the capital investments determine the cash flow.
• intellectual property costs.

Prepare a control and verification plan

develop a control and verification plan to prevent degradation (entropy) of the CTQs, high-level CTPs, and detailed CTPs in the detailed design over time(Git06, ch7.13, p246).

1 Documentation (for example, a flowchart or schematic) for the detailed design. 1 A list of the CTQs and CTPs. 1 The Voice of the Customer for each CTQ and CTP (nominal, USL, and LSL). 1 A sampling plan to collect data for each CTQ and CTP. 1 A data analysis plan for each CTQ and CTP (for example, a control chart and dot plots). 1 Actions needed to ensure a smooth transition of the detailed design to the Operations and Marketing Departments. * Actions are created as a result of the answers to the following questions: * Is the design ready for Operations/Production? * Is the design ready for Sales? * Have all stakeholders bought into the design? * Have all CTQs and CTPs been established for the design? * Will process owners use the control and verification plan?