Design for Six Sigma - Six Sigma

The title of this chapter has a dual meaning; the author believes firstly that the vast majority of the quality characteristics of any product are determined during the design phase of the project and secondly for many organisations the quality that they achieve occurs by chance and not by design. The purpose is to take the reader through the life cycle of a product from identification of the need to operation in the field. Although these notes are written about the introduction of a new product to the market place, all of the concepts apply equally to a service.

Design for Six Sigma (DFSS, or Quality by Design) can be applied in two situations:

  • When designing a new product, service or process: By giving attention at this early stage to customer satisfaction and variability reduction breakthrough changes can be achieved.
  • When traditional Six Sigma cannot achieve suicient improvement: Due to the limitations of working with an existing process it may not be possible to reach acceptable Sigma levels. There is a commonly held view that Six Sigma can only get a process to 5 Sigma levels and to reach Six Sigma we need to apply DFSS. This is spurious (as can be seen from the example project in this book) but the general principle that DFSS can deliver higher Sigma levels holds because it delivers more leverage and more options as discussed below.


The design stage has much more leverage for improvement than the latter stages, as decisions made there have a disproportionate effect on design quality and customer value.

The cost of problem resolution

Figure shows the relative cost of solving a problem in different phases of the design process. It is derived from information gathered during aerospace projects, but has become widely accepted throughout industry. During the design phase potential problems with the design will be identified and usually the cost of resolving these changes will be small if they are identified early enough so that consequential changes are minimized. If a decision not to make a particular change is made for whatever reason, for every £1 that would have been spent in the design phase, £10 will have to be spent in the development phase if it decided that the change is necessary after all. If it is not until production that a decision is made to implement the same change, the cost is now £100 for every £1 that it would have cost in the design phase.

Finally, that same change would cost 1000 times more if a decision is made to correct items already delivered to the customers. Figure is drawn to scale and yet the representation of in - service problem resolution costs is conservative because these are the costs associated with product recalls, or if the item is subject to dealer service, a free modification at the next scheduled service. All of the problems that do not justify such drastic action are not resolved for countless customers who have to put up with the standard of the product as they purchased it. Consequently, there is likely to be a hidden cost of not resolving problems that customers subsequently have to live with, namely, the costs associated with dissatisfied customers who may not only fail to buy the same product again, but also Publicise their dissatisfaction to their friends.

Committed costs vs. actual spend (Adapted from Yang and El-Haik, 2003)

Figure shows a typical relationship between the committed costs and actual spend during a product development programme of military aerospace equipment, but the author believes that there would be little difference in the shape of these curves in most product sectors.

The important concept there is that for most programmes, the design phase is conducted with minimal actual expenditure. Design teams have tended to be small in the early stages of the programme and financial commitment to the project occurs at a relatively late stage, often after early prototyping has commenced. The problem with this approach to product development is that experience shows that about 80% of the future costs associated with the product are likely to be committed by the end of the design phase, the stage during which minimal resources are expended. During the development and production phases when actual expenditure ramps up sharply, there is a decreasing opportunity to make changes which significantly affect the outcome of the programme.


When we are improving an existing process certain of the parameters we could control are out of scope. We cannot, for example, reasonably change the shape of a car panel as this would not only be very expensive, but would be highly disruptive to production. If we begin to work on customer focused, variability minimized products and services at the design stage we have much more flexibility in terms of what we can control.

All rights reserved © 2020 Wisdom IT Services India Pvt. Ltd Protection Status

Six Sigma Topics