MEMS Sensors for Novel Functionality in Tomorrow’s Appliances
And enjoy the advantages of disruptive or sustaining innovation in your products.
Consider disruptive innovation.
According to Clayton Christensen, Harvard Business School, “disruptive innovation describes a process by which a product or service takes root initially in simple applications at the bottom of a market and then relentlessly moves up market, eventually displacing established competitors.”
Cellular phones disrupted fixed-line telephony; in turn, the first Apple iPhone disrupted the feature-phone, and opened the door for many (Samsung, HTC, etc.) fast followers in the smartphone market.
The business-model-as-usual in the appliance space typically follows a sustaining innovation model; continuously improving established items, while charging higher, or at least steady, prices, has long been seen as the best way for appliance suppliers to achieve the greatest profitability for their wares.
But breathing down their necks are the innovators; think Nest thermostats, Dyson vacuum cleaners, and Keurig single-cup brewing systems, entrant companies with novel products that addressed the simple applications of residential HVAC control, home cleanliness, and a good cup of joe and, in the process, displaced incumbents. (And made a pile of cash; Nest was purchased by Google for $3.2B in 2014.)
What are ways an established appliance manufacturer can get a competitive leg up in the highly competitive consumer appliance space, whether the manufacturers’ appliance product lifecycles are, by plan, either following the sustaining innovation model, or riding the disruptive innovation curve?
The answer can be MEMS sensors: tiny, cheap, and reliable devices capable of sensing their immediate environment and then sending information about pressure, inertial motion, sound, vibration, chemical composition, or other properties, to electronics to receive and interpret the data.
Your smartphone goes from portrait to landscape when you rotate the phone, thanks to MEMS sensors. And tells you how many steps you walked that day.
Imagine these MEMS-enabled use cases:
A single-serving hot beverage brewing system is outfitted with an embedded MEMS pressure sensor that determines the altitude at which the coffee is being brewed, based on barometric pressure measurements. Knowing altitude, the brewing cycle in Santa Fe, NM, (7,200 feet above sea level) is then optimized to produce the best possible cup of coffee in Santa Fe despite the boiling point of water there being only 199° F, in contrast to the 212° F a New Yorker using the same machine would experience.
A sonic toothbrush head, disposable, is outfitted with a MEMS pressure sensor that creates feedback useful for addressing oral health in general, or periodontal disease specifically, by aiding brushing technique (harder, softer, correct angle), brushing being something which is still somewhat of an art when practiced at home, rather than a science.
A tornado warning device, similar to a smart thermostat or smart CO monitor, uses MEMS pressure sensors to “listen” for the infrasonic pressure signature of local tornado activity and give reliable advance warning that a highly destructive tornado funnel is approaching. Using Internet of Things thinking, these devices are connected house-to-house across whole neighborhoods in Tornado Alley; Big Data analytics predicts the tornado path and time-to-touchdown based on data received from the tornado monitors, and then produces a house-by-house report advising to either shelter in place, or evacuate, with the evacuation route being optimized to feature the best path(s) to safety, with evacuation route directions then sent directly to a subscriber’s smart phone and car navigation system.
In the first and second cases, the hot beverage brewer and the sonic toothbrush head, adding the MEMS pressure sensor is a sustaining innovation; in the third case, that of the tornado monitor, an entirely new, innovative appliance is being launched to disrupt established tornado warning systems, such as they are.
In all three cases, it’s MEMS devices that add desirable features to enhance the value proposition for consumers and it’s MEMS devices that add competitive advantage for the appliance makers.
MEMS is an acronym for “Micro-Electro-Mechanical Systems.” The term originally described the first silicon chips designed and constructed to simultaneously incorporate mechanical as well as electronic elements. The acronym continues to be applied to microchips today that incorporate elements of many different phenomena, including thermal, fluidic, chemical, optical, and even nuclear. Today, the unifying characteristic of these varied devices is that they are all extremely miniaturized (< 1 mm x 1 mm).
One of the manufacturing benefits of miniaturization, as exemplified by the semiconductor device industry, is that many devices can be made in parallel on a common substrate (e.g. silicon wafers), thus dividing the fabrication cost roughly by the number of individual devices per substrate, which can be in the hundreds or thousands of devices. Because many devices can be made simultaneously, MEMS unit production can rapidly scale to millions per year.
To source MEMS sensors, there are three typical options available: commercial off-the-shelf (COTS) sensors; custom sensors; and semi-custom sensors.
COTS MEMS pressure sensors are, by definition, stocked and available for purchase today. As a result, this option has zero development time and cost. As a further result, updated or new appliances employing COTS sensors can win the time-to-market race.
While COTS devices have zero development cost and time, these chips cannot be optimized for any particular application; their characteristics are by definition fixed. As a result, variables such as overall system cost may not be minimized, while other variables, for example reliability, are optimized.
How does this work in practice? After procuring a MEMS sensor, the next step in the design process is to integrate the sensor into the overall system. Typically the next step in the signal chain from the MEMS sensor is signal amplification, filtering, and analog-to-digital conversion (ADC), usually done on a single Application Specific Integrated Circuit (ASIC) chip. To properly connect the MEMS sensor to the ASIC, typically a third component, a printed circuit board (PCB), needs to be introduced. Signals from MEMS sensor are routed to the ASIC through the PCB. Each chip is connected to the PCB through eight wirebonds total, in this example.
In contrast, consider a simple example of a MEMS device where only the bondpads are customized for the application. The sensor’s pads can be directly mapped to the ASICs, and thereby eliminate one component, the PCB, and also eliminate four wirebonds. Since wirebonds are a typical point of failure in electronic systems, reducing their number simultaneously improves reliability and lowers cost. Removal of the PCB also lowers cost, as well as part count and overall size.
Here, the benefits of reduced component cost and reduced component count, and the concurrent increase in reliability, requires a sensor specifically customized for this particular application.
For a custom MEMS chip, typical development times and costs are on the order of years and millions of dollars. If the business plan calls for large volume applications, investments of this magnitude may be justifiable. Often custom development of a MEMS sensor is pursued for strategic considerations; there is often intellectual property that is generated in the development of the custom MEMS sensor, IP that can be used to differentiate system features and thereby gain competitive advantage. For novel sensors, this can provide a durable and protected advantage in the marketplace.
Semi-custom chips lie between the COTS and custom sourcing options. Semi-custom chips allow a degree of customization, but are based on an established fabrication method, an important advantage that allows for much of the cost and risk of development to be avoided. The development costs and times for semi-custom MEMS sensor sourcing are often an order of magnitude lower than custom, resulting in investments of time and money in months and hundreds of thousands of dollars, respectively, rather than years and millions of dollars.
Pursuing the semi-custom route with MEMS sensors can achieve cost, count, and reliability improvements, which can pay for themselves quickly, and achieve excellent time-to-market results.
Among these three options, choosing the “best” source for MEMS sensors often depends not just on the application for which they will be used, but also on the business plan and strategy as well. For example, for a company producing the coffee brewing system with altitude sensing, achieving the quickest and lowest risk path to market may be the highest priority. In this case, COTS sensors have adequate performance to address the need, barometric pressure sensing, and are the quickest go-to-market option.
For a disposable toothbrush head that can monitor the brushing pressure of the bristles for improved periodontal health, the unit volumes may be very high, and the per unit costs needs to be extremely low. In this case, the semi-custom sourcing approach may be well suited. The semi-custom sensor can minimize costs (such as fewer wirebonds, no PCB, smaller and cheaper chip) while also keeping development costs and times low.
Finally, a tornado sensor may benefit from being optimized to detect the infrasonic signature of approaching tornadoes. A custom MEMS sensor can provide the best performance for this critical application, as well as generate intellectual property to protect the device from competitors employing fast-follower strategies.