Specialize, Partner or Perish: Embrace Foundry Evolution for MEMS
By Jessica Gomez, Rogue Valley Microdevices
Jessica Gomez recently presented at MEMS World Summit (Munich, Germany November 11-12). On November 20, she became a board member of MANCEF.
Over the past decade the MEMS and sensors market has continued to build momentum, giving rise to a wide variety of devices ranging from gas and pressure sensors to ultrasound and specialized biomedical devices. The introduction of these tiny devices has enabled us to capture and use data in ways we never thought possible, fostering cross-disciplinary collaboration and driving innovation in agriculture, transportation, consumer electronics, healthcare and many other industries.
Startups, R&D groups, and even large technology platform companies are launching efforts to develop MEMS-enabled product lines, but in order to be successful, they must first understand the MEMS manufacturing ecosystem and how it is evolving to accommodate the coming influx of market demands.
MEMS Require Individualized Processes for Each Device
In the world of semiconductor manufacturing, it is routine for a fab to manufacture hundreds of different device designs using just a handful of process nodes. Semiconductor foundries share their design rules with customers, who then develop the mask set accordingly, literally adapting their designs to fit the pre-established and mature process nodes.
In stark contrast, most MEMS devices cannot conform to the level of standardized manufacturing processes that work so well for semiconductors. Rather, MEMS challenges us to develop individualized processes for each device. It’s one product, one process.
MEMS Foundries Require Customized Manufacturing Model
New MEMS designs generally emerge from either corporate R&D teams or academia, two groups that approach specialized MEMS foundries such as Rogue Valley Microdevices when they’re entering pilot or low-volume production. Today successful commercialization depends on open, accurate communication and close collaboration.
MEMS foundries must work side-by-side with product development teams to ensure that designs are based on real-world manufacturing process technologies. This highly customized manufacturing model makes it very difficult to support future demand for the groundswell of diverse MEMS devices that are in development. If we want to handle this upward trajectory of MEMS and sensors, we’ll need to adapt.
MEMS Foundries Moving Towards Specialization
While most existing MEMS foundries currently support a wide variety of devices types, I predict that market forces will cause our foundries to move toward specialization. Some companies will specialize in what they already do best, e.g., inertial sensors for the automotive industry.
Others might choose to develop their foundry business around a purpose-built facility, which, for example, only manufactures microfluidics or magnetic devices. Larger enterprises might opt to build captive foundries that are designed to serve their specific needs.
Creativity and Strategic Collaboration in MEMS and Sensors
Satisfying the thriving market for MEMS and sensors will require creativity. One idea: combine different disciplines of the manufacturing process at the same foundry. For example, we could have a biochemistry fab and a MEMS fab under the same roof, or we could have a MEMS fab and a packaging facility in one building. While these approaches may not yet exist outside of captive fabs or academia, necessity may drive them to fruition.
It will also require heightened strategic collaboration, a process that has already begun. To support both large volumes and greater diversity of devices, some MEMS foundries are building cooperative relationships with former competitors. Think of it as a restructuring of the supply chain.
MEMS Foundries and Supply Chain
Embracing the special challenges of MEMS manufacturing is worth our investment. We need to step back, individually and collectively, to understand where each of the existing MEMS foundries fits into the new supply chain so we can leverage our strengths. We can start this process by forging stronger alliances for tech transfer.
Once we begin to more freely share information, engage in joint product development projects, and engage with technology teams who are more connected and less guarded, we will finally begin to address the MEMS and sensor manufacturability challenges that are holding us back.
While we are unlikely to achieve the same level of standardization that has enabled the semiconductor industry to reach its great heights, as long as we evolve to meet demand, we will grow together and prosper.