Dave Fromm, Promex Chief Operating Officer and VP of Engineering.
Medical devices are getting smaller, smarter, and more powerful – much like the evolution of computers. In fact, many of today’s microelectronic-containing medical devices act very much like minicomputers -- sensing, seeing, gathering and sharing data, from both inside and outside the body.
But how best to approach the manufacturing and assembly of these increasingly complex medical devices? Ironically, by simplifying the steps. Robust yet elegant, a phased gate approach to metrology, measurement, and inspection of these products has proven effective. The results? Production of devices that feature tiny, thin materials that could quickly become troublesome, on time and within budget – all while ensuring quality.
Mirroring the product lifecycle process used to manage the development and sale of medical devices, Promex employs a five-phase gated approach to manufacturing microelectronic medical devices that revolve around the concept that the maturity of both the product design and its manufacturing process should be developed concurrently because how you build often influences what you build, especially for modern, highly integrated devices.
Explaining the five-phase gate approach
The phase gate approach is a best practice plan for numerous build iterations characterised by detailed measurements at all levels of integration, from basic assembly steps to top-level device function, with subsequent learnings guiding design and manufacturing improvements and changes along the way. It’s a process that decreases development risks while improving product quality, production yield, and time to market. It supports device development, from concept to production, and encompasses short prototype runs, ready processes, and large-volume production. Production teams work with engineering teams to make sure a conceptual design can be realised in production and assembly, as efficiently as possible.
Every one of the five phases calls for a series of deliverables that align with increased product maturity along the journey.
Promex Metrology inspection X-ray.
Phases 0 and 1
In early phases 0 and 1, multiple builds happen in small quantities, with 100% measurement of multiple parameters on every part. Investing the time and money needed to obtain up-front detailed measurements will create savings in the long run because changes can be simpler and more cost-effective when implemented on smaller builds. Critical outputs from these feasibility and development phases include discovering where the design affords good, high margins and where there are opportunities to improve the design, process, tools, and/or materials at low margins. These are the stages at which adjustments can be made to the top-level requirements of the product, such as determining if maximum performance is required if the device’s final cost can be increased to accommodate greater complexity, and yield requirements.
Most specs should be kept wide open at this stage, with the development team experimenting with any and all parts that have a shot at working together and within the design, as they determine the variations in top-level device performance and quality.
As a handful of builds are completed, measured, and accepted, the team can definitively identify what parameters are critical to device quality (CTQ) and function (CTF).
Phases 1 and 2
During the middle phases (late Phase 1 and Phase 2), medium-sized builds and more stringent acceptance of those builds start, with 100% measurements of CTQ/CTF parameters remaining and measurement sampling applied for higher margin and lower risk dimensions. Increasing the builds’ sample size will reveal intra-lot variations and multiple lots should also be run to show inter-lot variation.
At this stage, the team will commonly uncover new and potentially significant issues that had low occurrence and thus were missed by statistical chance in the smaller builds. Teams now have the opportunity to explore additional process improvements before solidifying the design and creating nearly final specs and tolerance requirements by linking the variations to top-level functions.
Phases 3 and 4
Moving into the later phases of 3 and 4, the team’s build sizes get large and there is increased sampling for measurement and fewer or no 100% measurements. Manufacturing functional tests and in-process QC measurements can now be used for product verification. This approach can save tremendous time and cost over a distinct verification process. Similarly, top-level functional characterisation can be used for product validation.
Now, specs can be refined until they become final. All parts meeting spec should be dubbed functionally good; those failing spec are functionally bad. Matching the two is a nuanced and complicated process because lower-level requirements seldom map one-to-one to top-level functional and/or quality requirements. When there are doubts between these connections, it’s best to establish an in-line functional test before moving the product through the manufacturing stage. If this is not feasible, specs should be biased to slightly over-reject good parts to help ensure that all released devices will perform per their intent.
Iteratively developed associated documentation will reflect the manufacturing build. This, along with measurements and test processes, comprise the true work product. If done correctly, phases 3 and 4 will yield the following:
- Development of appropriate inline quality control (what steps, tools, and specs carry the highest risks and subsequent controls).
- Linkage of low-level variation with top-level function.
- Real-time verification and validation, which is possible with careful development.
You will also gain an understanding of how variation at lower levels may impact the end product. For example:
- The absolute and relative positioning of components critical to function (CTF) and/or quality (CTQ) is especially important in manufacturing and assembling optical medical devices, when small alignment variations can adversely impact the product’s performance.
- Optically measured characteristics, sometimes combined with Scanning Electron Microscopy (SEM), for very small components, devices, or subassemblies may interact and cause downstream failure. For example, the flatness of a substrate upon different stages of build-up can cause warpage, and thus stress during manufacturing or (worse yet) in the field as a premature failure. Open connections occurring during assembly and CTE mismatches between materials bonded with a thermal cycling process can also cause reliability issues later.
Measuring with high-resolution X-rays at this stage can detect voiding within soldered adhesive interconnections. This is an especially important measurement for high-power or thermally sensitive products. Air pockets can lead to hot spots and rapidly degraded performance and/or product longevity.
Measuring with Confocal Scanning Acoustic Microscopy (CSAM) will detect voiding within epoxy fillers or moulded materials that can lead to delamination and other reliability issues upon temperature cycling or high moisture exposure.
Five-phase gate approach.
Taking measurements in all phases is key to success
Measurements during every phase are critical for manufacturing yield planning. Based on the information from the many available forms of measurement, manufacturing and assembly teams can create more precise volume and cost targets and improve their own processes while product owners can observe the quality and performance of their devices in the field.
A phase gate approach involves cross-functional teams focused on specific outcomes for each stage. Best practice documentation keeps the entire cycle on track. A very intentional, prescriptive list of steps can be an effective approach for bringing highly complex, conceptual ideas for the next best microelectronic medical device to life.