Anyone with a smartphone, tablet or digital camera recognizes that electronic devices are getting smaller. And, consumer delight with smaller electronics drives miniaturization trends in other industries.
Business users want products at work to be as easy to handle and operate as their personal devices. And high-volume consumer products tend to drive shifts in component availability and pricing.
Properly designed, smaller products can reduce cost. Quality may also be improved, since miniaturization requires higher levels of automation, leading to greater process consistency. However, there are also tradeoffs. Industries with low-volume legacy products may find the cost of redesign prohibitive, particularly in turnkey legacy systems designed with larger equipment footprints. Some companies are also reluctant to open the door to the greater design and manufacturability challenges that come with miniaturized products, until the competitive trends in their industry force a change in strategy.
Lean manufacturing principles and Six Sigma tools provide support for legacy and cutting-edge product strategy. EPIC Technologies’ customer base is primarily automotive, medical and industrial. For each of these industries, challenges must be addressed via different sets of Lean and Six Sigma techniques.
Automotive products are typically high volume. Advanced product quality planning techniques are used to optimize each design. The initial failure modes and effect analysis and control plan help identify likely design and process issues to address. Design verification, product validation and the production part approval process help identify and correct any remaining issues prior to process sign-off.
These assemblies typically have a high degree of miniature components and rarely incorporate through-hole components. PCB real estate utilization is high, and design for Six Sigma methodologies is used during PV to correct manufacturability or testability issues. Reviewed areas include critical to quality or special characteristics such as dimensions, tolerances, functional requirements and cosmetic appearance. Also, design for test outputs, particularly on manual assembly areas with high potential for low yields, scrap or escapes, are analyzed. The cost of an eventual escape could be extensive when ground and air transportation costs and penalties are considered. Corrections as a result of DfT analyses include streamlining of test times for some tests that are designed to be executed fully, even if steps have failed already. Modification to the test cycle is performed to immediately stop and print the ticket, without losing test time on an already defective board. Other improvements made are gradual and derived from the learning experience of relating test steps failed to components on the board. Reordering the steps also yields faster test times without sacrificing a robust diagnostic.
Automotive manufacturing processes are laid out in a tighter continuous flow scheme that follows Lean guidelines and supports a just-in-time philosophy. There is a focus on designing in high test coverage and use of complex functional testers. Continuous improvement focus incorporates Six Sigma statistical tools such as CPK, Gage R&R and control charts to achieve and maintain high first pass yield and low PPM defect rates. Design of experiments is performed to further improve process quality. The result is a high-performing process and quality product.
Comparatively, regulatory requirements in the medical industry drive a focus on consistent processes and minimal product or process change once a product receives regulatory approval. Medical device challenges can include lower volumes and higher product mix. As a result, there is more opportunity for waste elimination than in the automotive industry. Here, Lean techniques such as changeover time reduction, load leveling, line balancing and Kaizen events are used in the pursuit of continuous improvement. Product designs tend to require a bit more of the design for manufacturability and design for assembly techniques to improve manufacturability and assembly on the frontend, since design changes once a product has completed its regulatory approvals can be costly and time-consuming.
Often, required modifications identified through DfM analysis are selective step-ups or step-downs in stencils when a component requires a larger or smaller amount of solder paste applied due to special characteristics in shape or size. Modifying the entire stencil would affect other components negatively. Another challenge is pad sizes not appropriate to component sizes, particularly chip components. A good example of this would be 0603 pad sizes where a 0402 component will be placed. This mismatch most likely will generate tombstoning. Another common issue is caused by lack of coplanarity in shields. There is a bigger gap at the center of the shields than at the edges where they meet the board, and this creates a requirement for a larger amount of solder paste. Step-ups are not sufficient as a corrective action, and the use of an additional selective solder paste dispensing process is required.
The device inspection process can be complex, as many older products include odd-form components and wires requiring manual assembly processes. To improve inspection efficiency on these types of assemblies, EPIC uses inspection templates. Reduced size components are inspected with magnifying equipment such as 10x-20x mantises to detect failure modes, such as insufficient solder and proper placement during first piece inspection and sampling. These types of assemblies use more BGAs and shields, and less test coverage than automotive.
Additionally, some products are now classified as IPC Class 3. From an inspection standpoint, this forces the quality through inspection, rather than the manufacturing process, since mixed technology and manually assembled components are used. For example, one product has over 1,300 through-hole components and odd-form parts. IPC Class 3 has very stringent dimensional criteria, and applying these to an assembly with a high level of manual assembly requires detailed inspection. Use of Six Sigma tools to optimize the process is challenging because every assembly is different.
The industrial sector is characterized by a wide variety of designs. This requires creativity in implementing poka-yoke fixtures (the hold-down devices used to guarantee positioning/dimensions and ensure proper solderability), as some designs may not be robust to guarantee repeatability and manufacturability. Some designs require through-hole components and additional miscellaneous processes, such as conformal coating, PVA, epoxy applications, the use of OA solder paste, challenging line configurations and burn-in. Use of 0402 and 0201 component packaging drives the need for high-precision placement equipment with small form factor feeders.
From an EMS perspective, another challenging element is that many product transfers involve older product, which customers are reluctant to redesign. While these customers are open to using APQP techniques on new product, the vast majority of these projects involve older product. New product introduction cycles on these transferred products tend to be shorter than those found in automotive or medical products. Prototype cycles average five days. Flying probe testers are used at this point to shorten test and debug time.
This sector incorporates more complex box-builds than the medical and automotive sectors for EPIC, demanding more attention on the components/materials side. Some components are less robust because they have old packages, leaded or Pb-free with multiple alloy configurations. In some cases, components are only available through brokers. This can drive additional manufacturing processes or broader process windows to compensate for the amount of variation until superior alternates can be designed in. It also opens the door to quality issues driven by poor component storage procedures, counterfeit parts or simply component age. Process capability analyses are utilized to validate material content. MMC and cross-sections are performed as components get smaller and tolerances from manufacturers shorten.
As these examples illustrate, current trends toward smaller assemblies using highly automated production processes are not universally adopted across all industries. Legacy products present many opportunities for improvement through redesign. However, redesign costs tend to limit redesign activities in industries with longer product lifecycles. Lean philosophy and Six Sigma tools offer some options for reducing costs in these instances, but even greater benefits when applied during product development.
Carlos Rodriguez is a Lean Sigma Black Belt with EPIC Technologies; email@example.com.
Hilario Apodaca is EPIC Technologies’ director of quality; firstname.lastname@example.org.