Page 50 - EETimes Europe June 2021

50 EE|Times EUROPE

OPINION | MARKET & TECHNOLOGY TRENDS

Funding the MEMS timeline for a MEMS product development
Adequately scoping the funding and

Product Development effort requires first understanding the full
set of product requirements. A typical MEMS
product development begins with a bare-chip
Journey prototype made in a research laboratory.
While the functional requirements for the
MEMS device may be well-known to its
inventors, realizing a fully featured prod-
By Alissa M. Fitzgerald and Carolyn D. White, uct also requires an understanding of the
A.M. Fitzgerald & Associates customer’s interface and end-use require-
ments, all the way down to small details such
as suitability for wave soldering or tape-
OUR HOME STATE OF CALIFORNIA has wilderness areas of extreme climates, from desert and-reel delivery. Talking to prospective
to high-altitude, snow-capped mountains. Every year, people need rescue because they ventured customers and integrators early in MEMS
into the wilderness without proper equipment or training, often lacking the items necessary for development is essential to uncovering crit-
prolonged exposure to that environment, such as water or a warm parka. ical but non-obvious product requirements
Many development teams likewise begin microelectromechanical-system (MEMS) prod- for the MEMS as well as its electronics,
uct development — a challenging, multiyear journey — without enough of the most precious package, and software. These can inform
resource needed for success. In the desert, survival depends on having enough water; in technol- early design choices and thereby avoid costly
ogy development, success depends on having enough money. delays and redesigns much later in product
At our company, A.M. Fitzgerald & Associates LLC, we have completed more than 400 MEMS development.
development projects, about a quarter of them for MEMS startups. We have found that when new Once the full set of product requirements
products were unable to achieve commercialization, the most frequent reason was insufficient is understood, then the development work
funding, not issues with the technology or market. In collaboration with our clients, we usually may be more accurately scoped, and it
could build a MEMS prototype that functioned; but without enough funding for the long stages becomes possible to create a realistic devel-
ahead of scaling up manufacturing and commercialization — the period of prolonged exposure opment timeline and budget. For example,
in the MEMS development wilderness — the client’s product often perished. if the footprint of a proposed printed-circuit
Because money is usually a limited resource, it is important to understand the development board might be too large for a particular
timeline and to raise or budget enough funding from the start. The worst time to ask for more customer’s product, then a custom ASIC may
funding from investors or executives is when you’re almost out of money and the product is not need to be developed. When this require-
yet ready for sale. ment is known early on, the budget can be
The most common financial blunders we have seen made by both startups and established allocated up front for the ASIC; if discovered
companies during MEMS product development are: too late in the development, there may not
• Not reserving enough funding for developing the rest of the MEMS product, including pack- be enough funds remaining to make the
aging, electronics, and software switch, and then that customer opportunity
• Setting an unrealistic timeline for development, resulting in a cash flow problem would be lost.
• Securing only enough funding for the first run at a foundry Our experience has been that the
• Discontinuous funding, which causes wasteful loss of momentum engineering costs of developing a MEMS
product of medium complexity to the point
of validated foundry production (i.e., ready
for mass production and product sales) will
require, on average, four years and
US$4 million in engineering budget. (This
sum does not include the cost of business
administration, overhead, sales, or market-
ing.) It is certainly possible to spend much
more for more complex devices or product
systems. It is only possible to spend less
if existing IP, such as a foundry process
platform, is cleverly leveraged to accelerate
development.
Starting a MEMS development without
this timeline in mind and an adequate sum
of money lined up would be like walking
through the desert without enough water.
You might make it, but it will be a stressful
journey at best. In MEMS development, just
Typical engineering budget required to develop a MEMS product through four stages of as in wilderness adventuring, things rarely
development, to the point of volume production readiness. Does not include the cost of go exactly as planned: Wafers break, debug
1
business administration, overhead, sales, or marketing. (Reprinted with permission from takes longer than expected, customers
“MEMS Product Development: From Concept to Commercialization” [Springer, 2021]) change their minds, and random events like

JUNE 2021 | www.eetimes.eu