Design for Six Sigma

IBIS UvA approach

Regular Six Sigma projects are mainly conducted in the operational part of organisations. Stagnations and structural problems are tackled; improvements often are found in the form of a control system or modifications in the standard way of working. Occasionally a redesign of part of the process is needed.
A major part of the problems in processes can be prevented, however, by taking possible problems during manufacturing and operations into account during product and process development. In order to apply the basic principles of Six Sigma in product and process development, an adaptation of the methodology is required. This adapted methodology is called Design for Six Sigma (DfSS).

Essential elements of DfSS
An important part of DfSS is design for manufacturability: to design products and processes in such a way that they result in less problems during manufacturing. Standard principles are:
- robust design (design products and processes in such a way that they function well in non-ideal circumstances).
- reduce complexity of products and processes (thus reducing the probability of mistakes).
- inventorize as early as the design phase wich mistakes and problems are likely to occur, and design preventive mechanisms.
Altogether the emphasis is on robustness and mistake proneness, and less on ideal performance.

A second important idea in DfSS is a focus on stakeholders. Good product and process design is not technology driven, but is driven by what stakeholders consider to be value. In DfSS this is embodied by a disciplined translation process that starts from the stakeholder. His or her functional requirements are translated to technical requirements, and these into product specifications and process settings.

Finally, DfSS strives to achieve a more efficient design process. The main techniques for this purpose are:
- Critical parameter management: relations between process settings, product characteristics, technical requirements and functional requirements are systematically quantified. In this way, specifications and capabilities can be translated from higher levels (functional and technical requirements for the product) to lower levels (process settings) and vice versa.
- Early warnings: on designated moments in the design trajectory specialists in areas like manufacturing, reliability, marketing, maintainability, et cetera, give feedback on concept designs.

IDOV
Regular Six Sigma projects follow the DMAIC method (Define, Measure, Analyze, Improve and Control). For DfSS projects there is a modified methodology: IDOV. The phases are:
1. Identify: functional requirements for the new product; the translation to technical requirements; linking the design project to the strategy of the corporation.
2. Design: development of concept designs; selection of the most promising ones; identification of design parameters, but also nuisance variables and disturbances; feedback of various specialists on vulnerabilities and risks.
3. Optimize: based on the results of experiments the product specifications are established, as well as process settings.
4. Verify: test runs; design of a quality control system.

DfSS employs many techniques which are familiar in the regular Six Sigma programme. The main additions are: critical parameter management, TRIZ (theory of inventive problem solving) and reliability engineering.

Curriculum

Identify
Determine functional requirements
Translate into technical requirements, operational definitions
Create business case
Critical parameter management
Measurement system analysis and gauge R&R studies
Control charts and process capability analysis (incl. non-standard PCA)

Design
Develop concept designs
Establish high-level design, Pugh concept selection matrix
Identify control and nuisance variables
The design review
Idea generation: effective brainstorming, autopsies, TRIZ, pairwise comparison, and other approaches.
FMEA and mistake proofing
Statistical data analysis: ANOVA, regression, cross tabulations, logistic regression

Optimize
Determine CTQ-X relationships
Determine optimal settings
Determine tolerances
Design and analysis of experiments
Robust design
Tolerance design

Verify
Test run and capability analysis
Design quality control system
Project closure and transfer
Process improvement and control: process mapping, shopfloor management, planning for control
Lean manufacturing, incl. line balancing, value stream mapping, rapid changeover, 5S
Statistical process control, PID controllers, and other control systems
DMAIC overview
Analysis of unplanned data 

More information, Holland Innovative / Dommelvalley Platform.