[Please note that the following article — while it has been updated from our newsletter archives — may not reflect the latest software interface and plot graphics, but the original methodology and analysis steps remain applicable.]
Guest Submission - Lisa Allan, Delphi Thermal Systems
With the ever increasing pressure to improve quality, reliability and reduce warranty costs, companies are utilizing many different problem solving tools and methodologies such as Six Sigma, Design for Six Sigma (DfSS), Taguchi's Robust Engineering, Shainin's Red X, etc. As a result, companies are spending a significant amount of time preventing the recurrence of problems. However, the real gains are in taking the next step: determining preventative action prior to the problem actually occurring. This proactive problem prevention concept is what the Japanese call Mizenboushi .
The need for product development teams to foresee a problem and prevent it from happening led the Thermal Systems division at Delphi Corporation to adapt the philosophy of Mizenboushi. Since the concept of problem prevention is the cornerstone of reliability, we chose to lead our process as part of the Reliability Engineering group. Two of the critical tools we have developed that are the main topic of this article are Change Point Analysis and Design Review Based on Failure Mode (DRBFM).
The first key tool is Change Point Analysis, which has the following objectives:
The second key tool is the proactive problem prevention method that is called DRBFM . This tool helps find problems through a forum of Good Discussion with a cross-functional team.
The left side of Figure 1 shows how to move your organization from a continuous improvement (fire fighting) culture to a problem prevention culture. It is important to begin with a rock solid foundation that includes stable, robust product and process designs, built on the standard work of design guides, standard test procedures and Failure Mode and Effects Analysis (FMEA). If you do not have stable and in-control designs with standard engineering work, you will end up with too many changes to the product which won't allow the engineering team to adequately deep dive them all. Toyota spent many years establishing the foundation of high quality and reliable products for their customers before deciding to go further and attempt to predict problems prior to their occurrence.
Figure 1: Problem Prevention Foundation
The problem prevention process consists of three action-based steps: Good Design, Good Discussion and Good Dissection. The GD3 methodology is represented in Figure 2 below .
Figure 2: Problem Prevention Process
Good Design means you have stable, robust products with a good FMEA. For Good Discussion, Change Point Analysis and DRBFM are the key tools used by the product teams to drive the deep and thorough discussion, concentrating on change points and focusing on the weak links within the design. Because the review begins with a good design, the team knows that their risk lies in where they make changes. Once designs have been discussed and parts are built for evaluation you move to the final step in the GD3 methodology, Good Dissection. This step contains the Design Review Based on Test Results (DRBTR) and Design Review Based on Design and Process (DRBD&P). The key for Good Dissection is for engineers to look at and compare prototype or production design intent parts to results from test and build, always looking for changes from the ideal state.
Delphi Thermal has developed a road map (Figure 3) to aid us in executing the GD3 process. This article will discuss in more detail each of the five steps individually.
Figure 3: Delphi Thermal Road Map
What constitutes a change point? You need to remember where change can occur, in design, supplier, usage environment, customer or internal specifications, etc. This is where the Discovery Checklist (Figure 4) is used. Delphi Thermal uses this checklist as a brainstorming tool to aid in determining what the true changes are. The beginning of problems lie where changes have been made, so if we pay close attention to and manage the changes, we will have Good Designs.
Figure 4: Discovery Checklist
In Figure 5, we look at the ability to dig deep into these changes. When the number of changes to a carry-over design have been minimized, the product development team is able to increase the depth of their discussion. For instance, if you make 20 changes on your design, you will not be able to adequately dig deep into each one of those changes. However, if you make three changes to your design, you will have the ability to truly deep dive and thoroughly discuss each change. Again, being able to minimize the number of changes gets you back to the original premise that you are starting from a stable, robust design with a rock solid foundation.
Figure 5: Minimizing Change
At Delphi Thermal, Change Point Analysis is an initiative with a couple of goals. By stressing the importance of the Change Point Analysis first, we are identifying our areas of greatest risk. Secondly, the process helps to focus engineering personnel, the use of resources and enables the prioritization of changes with all potential concerns. Forecasting these problems begins with the skills of a Reliability Engineer who helps facilitate team workshops by using reliability tools to predict when and where problems will happen. The prevention of reliability issues is now possible because we have made any potential problems visible.
To begin the process, Delphi Thermal starts by asking the lead Product Engineer to list the changes in the Change Point Summary Sheet (shown in Figure 6). This involves picking a baseline, comparing to the baseline and really understanding what is different, documenting concerns/impacts and identifying the cross-functional team members. To help with this activity, the Product Engineer can utilize the Discovery Checklist (Figure 4). We have found this up-front work, prior to the Change Point Analysis workshop, to be more effective than bringing a blank sheet to the team as engineers prefer to critique at items people have missed or incorrectly identified. This creates that forum of Good Discussion. This pre-work allows the Reliability Engineer to lead the team into the Change Point Analysis workshop with the identified cross-functional team, which includes subject matter experts. Items that we ask the engineering team to bring with them consist of drawings, parts, specifications, etc. for both the baseline design and the proposed new design. During the workshop, we have found that assigning a risk level, prioritizing and defining a risk reduction strategy has worked well for us. The working level team will take this risk assessment to leadership for their approval and understanding.
Figure 6: Change Point Summary Sheet
The purpose of the Change Point Summary Sheet is to identify the baseline design and detail changes off of this selection. The Product Engineer should utilize the Discovery Checklist to help avoid missing any changes. The Product Engineer must also consider how changes relate and impact each other. For example, the engineer needs to consider how a dimensional change potentially interacts with a material change. In the concerns/impacts column, list all potential failure modes, areas of risk and their impact. The last few columns of this worksheet go after the risk priority assessment and justification. Delphi Thermal determines what strategy we will utilize to mitigate the risk for all line items. For example, some risk reduction strategies we might consider are Finite Element Analysis (FEA), routine engineering validation, launching a DfSS project, DRBFM or FMEA. Those highest priority DRBFM risk reduction strategy items, based on changes from the standard baseline design, are then transferred to the DRBFM form. We have found prioritization necessary as we generally make multiple changes and this helps to truly deep dive those items most at risk.
Before getting into the DRBFM, I would like to point out that we have encountered team members wondering how this process fits into the DFMEA process as the two appear quite similar. At Delphi Thermal, we have incorporated both tools into our risk reduction strategy (see Figure 7). Our customers still require DFMEAs and in some cases now DRBFMs. Both tools are applicable during the product development process but the DRBFM enables us to dig deeper on certain high risk items. In Case 1, for a totally new design/product, with no current design guide or FMEA, we begin with a DFMEA, which helps us to create those standards. We work on building that solid foundation mentioned previously. In Case 2, for a next generation product or a carry-over design, we begin with a DRBFM and update the design guides, FMEAs or standard work to help create that new baseline design.
Figure 7: DFMEA or DRBFM Assessment
DRBFM is the second step on the Delphi Thermal road map (shown in Figure 3). Here the focus is now on those vital few changes. The change object is identified and a functional analysis around this object is created. The change object is one level above the change point or the part name. For example, change points would be painted versus unpainted, bent versus straight and the change object would then be the heat exchanger frame. The functional analysis would be done on the frame.
The DRBFM worksheet is a living, working document that has three separate sections, as shown in Figure 8. The first section is a detailed FMEA on identified change points. The second section is the results section. Results can be detailed around the following:
Figure 8: DRBFM Worksheet with three separate sections
The third section is for management and control. Some tips that we have learned in doing this DRBFM process are:
The immediate output of the Change Point Analysis and the DRBFM is to make your designs and manufacturing processes better. The worksheets capture all the good work and learning. Then at the end of the process, the worksheets can go away as the learning has been documented in design, process or test standards. The acquired knowledge must be shared across the organization so that others will be able to appreciate your discovery. (For those with Japanese customers, the concept of sharing your lessons learned is called Yokoten.)
Going back to the Delphi Thermal road map (see Figure 3), the third step in the process is Design Review Based on Test Results (DRBTR). This is the examination of parts after test or a tear down analysis. The key to this step is to accomplish Good Dissection. Parts need to be closely examined after testing and it is very helpful to examine the parts while comparing to a reference part that was not tested. In addition, take measurements with simple measurement tools such as calipers, micrometers, rulers, etc.
The fourth step in the process is Design Review Based on Design and Process (DRBD&P). This is the final design review comparing parts produced off of production tooling to the parts made during prototype. During this step it is imperative to fully define the detailed requirements of what yields a good design.
The fifth and final step in the process is the Closure phase. Delphi Thermal focuses on two key elements in this phase:
At this point, we close out the problem prevention activities for this particular customer program or design.
Figure 9 depicts our typical development process and the timing when the problem prevention activities take place. Delphi Thermal looks to begin the Change Point Analysis workshop right at the end of pursuit and award of business. At this time everything comes together to understand change points and risk around the design.
Figure 9: Development Process Timing
The key lessons learned from this process are:
In summary, the product development team needs to focus on a detailed Change Point Analysis prior to beginning the DRBFM. In following this path, an organization will be able to concentrate their efforts on those truly high risk items, which in turn will allow the team to foresee potential problems and prevent them from occurring. In addition, standardization of designs will take place, which supports the building blocks of our foundation, including stable, robust product and process designs.
 Yoshimura, Tatsuhiko, Toyota Style Mizenboushi (Preventative Measures) Method GD3, How to Prevent a Problem Before It Occurs, JUSE Press Ltd., 2002.
Lisa Allan is currently working as a Sr. Reliability Engineer at Delphi Thermal Systems. She joined Delphi in 1989. She holds a B.S. and M.S. in Industrial Engineering from Alfred University. She is Delphi certified as a Design for Six Sigma Black Belt as well as a DRBFM Expert. She can be reached at [email protected] for additional information.