"Talent wins games, but teamwork and intelligence wins championships." –Michael Jordan
Hundreds of PCB designs are released to be quoted or fabricated every day around the world. Most of the released designs will have engineering questions (EQs), which are also called technical queries (TQs) in Asia, generated once the data package has been received and analyzed. This is the accepted data transfer methodology, as it has been even before disco music was invented (and thankfully disappeared).
CAD software tools and their associated DFM software packages have not been able to resolve this problem by getting in sync with fabricator and assembler DFM/quoting software and rules. In part, this is due to fabricators not providing their design rules due to IP concerns. It is also due, in part, to the common practice of sending non-intelligent data along with conflicting e-paper documentation (e.g., fabrication prints).
Until the design-to-fabrication design transfer environment makes a revolutionary leap, there is a manual method that can reduce data transfer issues. As Michael Jordan’s quote references, we have lots of talent on both sides. We have teamwork resolving the issues. Let’s focus on the intelligence aspect to create a championship.
Establishing the methodology is the easy part, so how can this be performed? In this column, I’ll outline how to use these seven fundamental steps based on Lean/Six Sigma concepts:
- Establish the problem.
- Establish the goal.
- Collect and analyze data.
- Improve the process.
- Update procedures and training.
- Declare victory.
- Go back to Step #2, update the goal, and keep the project alive until no EQs are generated.
The problem is there are EQs being generated for all or some designs that are received by the fabricator. Both sides of the transfer need to agree that this is an important problem, or there is no need to proceed. My experience has shown that it is most successful when this problem is presented by the fabricator to a customer.
Create two KPI goals to reduce the number of EQs and EQ communication cycles by 50% within six months. This is most successful when the KPIs are added to both the design team and fabricator front-end engineering team’s performance goals. This puts some skin into the game at both ends.
The fabricator should review all the designs received over the previous six months. Log the part numbers, the design center locations (if known), the number of EQs generated for each part number, each revision of that part number (include any pre-design DFMs that were performed), and the number of EQ comment back-and-forth cycles. Then, create a spreadsheet with each EQ issue. Pareto out the EQs by part number and revision and the EQ quantity by type.
The customer can take the data and Pareto it out by the designer, project, design location, etc. They can also collect data from all their suppliers to compare the EQs that were or were not noted. For example, part number Z is built by a local quick-turn shop and then transferred to high volume shops in Asia. There may be issues not noted by the QTA shop that the volume shops note.
Don’t try to solve every issue at once. Apply the 20/80 rule. For most of these analyses, 20% of the problems will generate 80% of the total EQ quantity. Update customer fabrication/acceptance specifications and design electrical/mechanical requirements that can be adjusted. The customer can review where the design conflicts with their acceptability specification, review data output translation issues, etc. Fabricators should review their quality data to see where they have sufficient capability margin to allow a wider design rule to be used for EQs where the customer states that the request cannot be accepted. This is where the fabricator provides more knowledge to the customer about their process capability than prior communications.
A global deviation document can then be generated, stating where both sides agree on changes to their existing procedures or capabilities. For example, approval for non-functional pad removal can be included in this document. The best solution is where the OEM can create one document that can be applied to all suppliers to keep it simpler for the designer. Items in this document will not require an EQ for future designs since they will have global approval.
Concurrence at this point is where the intelligence is created to win this championship. You have now achieved Step #6, where the team declares victory.
The fabricator can monitor the Pareto results for design transfers over the next six months. Generate a new analysis and present it to the customer. If the 50% reduction has been achieved, reset the level down another 50% and continue the project until 80% of the design transfers achieve the zero EQ goal.
Teams that I have been involved in have successfully run this project with multiple large and small companies around the world. The design team becomes more efficient because they do not have to stop work on their new design to address issues on the prior design. The fabricator gains substantial net capacity because the designs flow through planning/CAM and get to the production floor faster. Overall, this manual EQ reduction method will create championships by reducing design costs and attract additional revenue for the fabricator because it will be easier to work with you versus a competitor.
Dana Korf is the principal consultant at Korf Consultancy LLC.