In modern computing and electronics, the transistor tends to get the credit – and rightly so.
It is the foundational switch of the digital age which every processor, memory chip, and digital system ultimately depends on.
CEO and co-founder of SonicEdge Ltd.
But running alongside it, built from the same silicon, is a parallel technology with far less attention, but follows the same logic: if you can carve a transistor into silicon, you can carve anything.
That field is Microelectromechanical Systems – MEMS.
Silicon as a mechanical material
The defining mindset shift regarding silicon’s capabilities came in 1982, when Kurt Petersen published “Silicon as a Mechanical Material” – until this point, silicon had been treated almost exclusively as an electric material. Petersen’s idea was deceptively simple: the same photolithographic processes used to manufacture transistors can also be used to produce microscopic mechanical structures that flex, pump, resonate, and deflect.
This shift created a new category: MEMS. Instead of assembling mechanical parts, engineers could now fabricate them using semiconductor processes – alongside the electronics that interpret their signals. However, MEMS is not exclusively a silicon technology.
Devices can also be built from glass, piezoelectric ceramics, polymers, and compound semiconductors, depending on individual physics demands, but silicon dominates the space as it sits inside a global manufacturing ecosystem – with a scalable and cost-effective supply chain already in place.
The first commercial wave
MEMS established its value – outside a laboratory setting – in applications where size and precision were critical for successful operation.
The inkjet printhead came first: microscopic nozzles, chambers, and heating elements ejecting droplets of ink with sub-millimeter precision. With the application of MEMS technology, the device’s previous, complex mechanical systems became a solid-state fluidic device manufactured like a chip.
MEMS’ second breakthrough came in automotive safety, in the form of the airbag accelerometer. Before MEMS, crash detection relied on mechanical systems such as the rolamite – a roller inside a tensioned band. This too was mechanically complex, and consequently sensitive to wear and difficult to safely mass-produce.
In 1991, Analog Devices introduced a MEMS chip – a tiny, suspended mass that shifts under sudden acceleration, and whose signal can trigger airbag deployment in milliseconds when processed electronically. This alternative was smaller, faster, more reliable, and far easier to manufacture at scale, causing MEMS-based sensors to rapidly replace mechanical crash detection systems in mass-market vehicles.
The invisible infrastructure
Once MEMS proved reliable at scale, the technology began to spread into the core of modern electronics.
A defining example is the Digital Micromirror Device (DMD), developed by Larry Hornbeck at Texas Instruments. The device consists of millions of microscopic mirrors, each individually steerable by electrostatic force, switching at kilohertz rates. Each mirror represents a controllable pixel of light, and is the foundation of Digital Light Processing (DLP) projection – which is used in many cinema screens and office projectors today. The technology is invisible to the end user but essential to the system.
Today, MEMS technology underpins many modern electronics, quickly outpacing preceding legacy architectures:
1. Motion sensing in smartphones and wearables – i.e. gyroscopes and accelerometers
2. RF filters and band-switching in mobile communications
3. Oscillators replacing quartz crystals in compact timing devices
4. Microfluidic chips running chemistry at the cell scale
5. Optical MEMS switches routing light between fibers inside AI data centers
The manufacturing foundation enabling this scale is Franz Laermer’s deep reactive ion etching process, developed at Bosch in the 1990s. The technique allows precise, high-aspect-ratio silicon structures to be produced at an industrial scale – an essential for modern MEMS production workflows.
Microphones: the pattern
While the systems described above are built with MEMS at their core, the technology first became visible to consumers at scale through microphones.
MEMS microphones were introduced in the early 2000s, offering something its predecessors could not: wafer-scale uniformity and consistent units from a single production run, at dimensions that no preceding microphone technology could approach. This consistency changed system design.
At the time, mobile phones typically had one microphone. Today, they often have several. Likewise, earbuds contain multiple microphones, modern vehicles integrate upwards of eight, and emerging next-generation devices – such as smart glasses – are moving toward even larger amounts. This integration is to run capabilities such as superior beamforming, active noise cancellation, and spatial voice isolation simultaneously, and is driven by economics.
MEMS made each additional microphone cheap enough that adding it became a logical next step; product designers now move from ‘one best microphone’ to distributed sensing architectures.
MEMS Speakers: rethinking how sound is generated
If microphones were an early success, then speakers were the next MEMS frontier – and the hardest problem to solve.
This came down to how sound is generated. Audible sound requires moving meaningful volumes of air, and moving air demands physical displacement at a scale that stiff silicon structures are not optimized for. Every attempt to transpose the conventional speaker model into a MEMS form factor ran into the same constraint: a rigid microfabricated membrane cannot move enough air.
The solution required letting go of the conventional model entirely.
A conventional speaker membrane pushes air hundreds of times per second, and must be large enough to do so. MEMS architecture replaces that membrane with an ultrasonic air pump – a small, stiff membrane paired with an acoustic valve, together cycling hundreds of thousands of times per second – replacing size with speed.
What the membrane loses in displacement, it recovers in frequency, delivering the same acoustic output from a structure a fraction of the size. Silicon’s stiffness now becomes an advantage; it enables precise, stable, high-frequency operation without mechanical distortion or fatigue.
The result is a speaker delivering full-range audio from a chip-scale component, manufactured using standard MEMS processes. This unlocks new form factors and device capabilities, including: ultra-compact in-ear devices, invisible audio systems, and assistive technologies that look more like sleek accessories than bulky health devices.
With MEMS speakers, audio is no longer constrained by diaphragm size, and devices can now be defined by the function they deliver – not the capabilities limited by their size.
Plenty of room
MEMS technology has shown a consistent pattern: when mechanical systems are reimagined at microscopic scale, entire product categories change.
For example, Larry Hornbeck’s micromirror arrays became the foundation of modern projection systems. Accelerometers moved from bulky mechanical assemblies into invisible chips inside smartphones.
Today, MEMS continues to expand into new domains, including advanced sensing, optical routing, medical systems, and hybrid electromechanical architectures.
What began as an extension of semiconductor fabrication is now becoming its own design paradigm – unifying the separate disciplines of mechanics and electronics.
We feature the best business computers.
This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit
https://cdn.mos.cms.futurecdn.net/gMavKmnr8ZGuzV56Abc7DM-2560-80.jpg
Source link
