Martin Linder columns | December 12, 2005

DfX - Design for Excellence

DfX, or "Design for Excellence", is a major concept that presupposes that aspects such as producibility, testability and/or overall cost structure are included early in the product's development phase.
For small and medium-sized companies with products whose manufacture is placed out with partners, problems arise as they do not themselves normally possess the required expertise as regards production. The key to success is co-operation. This article provides suggestions for how a joint project should be structured and planned. There is a brief presentation of a method that guarantees that all parts of the product will be taken into consideration.

In recent years, those of us who regularly read the electronic press have become used to seeing abbreviations along the lines of "DfM", "DfA" and "DfT". There has been an abundance of more or less creative abbreviations of this type. They stand for various "Design for " activities, such as "Design for Manufacturing" and "Design for Assembly". The fact that they have started to turn up so frequently in the press is a symptom of how important it has become to take production aspects into consideration early in the product's development stage in order to keep down overall costs.

With the short market windows we are now witnessing on the global electronics market, it is extremely important for the product to be correct from the outset. This fact, along with the short "time to market", places demands for an efficient product development process. The question we are faced with is how to incorporate these aspects successfully into the projects. The difficulty is often that the knowledge is possessed by individuals other than those who are normally based in the R&D department of a small or medium-sized product owner company, where production is performed by an external contract manufacturer. The answer is co-operation.

DfX, or "Design for eXcellence", is a major concept that incorporates "Design for Manufacturing", "Design for Test", "Design for Quality", "Design for Cost" and all other design aspects that you can come up with. This breadth is what makes the concept difficult to handle, yet at the same time makes it powerful. By looking at various "Design for" projects, it is possible to identify a large number of general truths that can be utilised to find an efficient and successful working method.

We can start by observing that the lowest common denominator for these various types of design project is that they focus on one product or one product family. Furthermore, the projects tend to have such a large work content and a relatively high level of complexity that the required knowledge seldom or never exists in a single person. It is therefore necessary to put together a project group, and the knowledge and the experiences that are required in this group have to be carefully considered before the start of the project. The work within the project group must be structured in such a way that no aspects of the product are ignored. It is also very important to ensure that you do not end up in a state of sub-optimisations, rather that the group keeps the global objective in sight throughout the entire project.

The product may be in various stages of its life cycle, such as the development phase, the prototype phase, the pre-serial production phase or the serial production phase. This is important, as it affects the degree of latitude for change that exists in the project. It is naturally simpler and cheaper to introduce an essential design alteration in the development and prototype phases than in the mass production phase.

The project group
The product owner possesses extensive knowledge about the product's design, history and market. When it comes to production, testing and quality, however, the knowledge is held within the producer's walls. It is important to bear this in mind when putting together the project group. We can verify that the success of the DfX project is important, both for the product owner and the producer. Consequently, both parties have a strong incentive to participate actively in the project.

Here is an example of a successful project group for conducting a DfX project focusing on lowering the overall cost structure for a product within the industrial electronics segment.

Organisation of a successful DfX group

The group comprised a project manager, a production expert, a person who participated in the discussion regarding alternative building methods, a person who had long experience of production in the industrial electronics segment, a person who had been involved throughout the product's history, one of the designers and the product manager, who had a very good insight into the market's wishes and requirements.

One very obvious risk for the project is that it results in a number of changes, each of which may appear positive, but when taken together do not benefit the project's global objectives. The easiest way of insuring yourself against such 'sub-optimisations' is to have one person whose role in the project is constantly to maintain the helicopter perspective. This person should ideally be well-acquainted with the product, and should play or have previously played a leading role as regards the development or production of the product.

The project
The success of this project is largely dependent on the combined creativity within the group. A good way of encouraging creative thinking among the participants it to launch the project with a meeting for open discussions focusing on the product. This enables the project to get off the ground quickly. To ensure that it does not go completely off the rails, it is a good idea for the project manager to make sure the meeting has some kind of structure. The method that has proved most successful is to focus on the product (an actual example), but at the same time systematically to work through all the constituent parts on the basis of the BOM list (Bill Of Material). Working from the BOM list, it is possible to establish a tool whereby the group can note which measures have to be taken for each of the components and who is responsible for these measures being carried out. This method guarantees that the group will not miss any part or component of the product. The difficulty is not to get caught up in sub-optimisations, but rather to keep the overall objective of the project in sight at all times.

Tool for DfX project in the form of an Excel worksheet based on the product's BOM list

As previously described, the projects are launched with a start-up meeting at which a multitude of ideas and possible changes will emerge. The next stage is to analyse the consequences that the changes will entail, both as regards the current design parameters (producibility, testing, costs, etc.) and as regards how feasible it is to implement the change. This can be discussed at the start-up meeting if there is time, otherwise it is a task that each participant can carry out him/herself. We refer to this phase as a "feasibility evaluation".

The various stages in the project

After this it is time to hold another meeting under similar circumstances to the first meeting. The meeting should encourage open discussions. During this meeting, which we refer to as "Selection", we decide on the proposed changes with which we intend to proceed. It is also important to provide latitude to bring up additional ideas. This has proven to be extremely valuable, as everyone has now had the opportunity to consider matters for themselves. It has often been the case that the ideas that have borne most fruit have emerged at this point in the project.

The subsequent work relates to compiling the changes that will be implemented, as well as ascertaining why and how they should take place. This includes producing data for complete tender inquiries, as well as highlighting any tool costs or other investments that may be necessary.

After the concluding summary meeting, the engineering work itself will begin in the form of actually introducing all the changes and verifying that they are having the desired effects in production.

We have previously made the judgement that as DfX projects of this size require so much experience within a wide range of different disciplines, it is only now that we at NOTE feel secure in offering our customers this service across a broad front. This is because we have grown so strongly in recent years that we now have more than 10,000 man-hours of experience to share.

The fact that this is a new service means in turn that it is difficult to demonstrate its effects and strengths statistically. However, we have conducted an internal project on one of our customers' products in order to teach ourselves how to structure the DfX work.

The product in question is a plastic box containing control electronics. It sounds simple, but it generates a large amount of heat and has stringent demands as regards its environmental durability. In addition, the product comprises power electronics, which means that a considerable amount of wiring has to be fashioned and secured. The result of the project was that we succeeded in reducing the number of mechanical components from 17 to 11, the test time from 48 to 22 seconds, the manual production time from approximately 5 to just over 3 minutes, and with reductions in materials this resulted in a cost reduction for the customer of around 40%.
Investments in the form of new tools, hardware for testing and fixtures for more efficient production amounted to approximately SEK 1,500 thousand. Working on the basis of the customer's forecasts for 2006, it can be seen that the DfX project, including subsequent required investments, paid for itself in a couple of months.

Martin Linder
Load more news
December 12 2018 12:21 pm V11.10.9-1