MEMS & Sensors Industry Group

BackgroundSEMI’s Device Working Group, part of its MEMS Sensors Industry Group (MSIG), actively works to lower barriers for MEMS-based gas sensor market adoption. In 2022, the group published SEMI MS14 – Guide for Critical Parameters of Gas Sensors [1, 2], which recommends guidelines for gas sensor product datasheets to improve standardization and adoption.This article explores what gas sensors are, how they are calibrated, what their limitations are, and how the next generation of MEMS-based gas sensors is driving innovation. Here, we utilize terminology that the most common, accepted, and recognizable across different sensing communities.From Gas Sensing Elements to Gas Detector SystemsFigure 1 highlights the primary components related to gas sensing. The most fundamental component, the gas sensing element, responds to changes in gas concentration. It is an analog circuit without additional electronics, and it often has only one output like resistance, current, light intensity, or voltage. It is sometimes called a gas sensor, but the recommended terminology is gas sensing element. Such terminology is the most popular, accepted, and recognizable across different communities.Figure 1: Components related to a gas sensorThe gas sensing element can connect to an additional electronic circuit that includes a signal conditioning module, an analog-to-digital converter, and as of recently, an onboard edge data processor. MEMS microfabrication technology allows multiple gas sensing elements and electronic circuit components to be fabricated as an integrated circuit module. These elements can be as small as a grain of rice. This combination of elements is often called a gas sensor, gas sensor system, or gas sensor module. However, the recommended terminology is gas sensing module.The gas sensing module can connect to additional electronic components such as power management, communication, human-machine interfaces, and others. It can also be supplemented with software or firmware and packaged in a mechanical enclosure to create a complete gas sensing instrument. Such gas monitors can be designed as stand-alone, stationary, handheld, or wearable systems, or embedded into bigger systems. The gas sensing instrument is commonly referred to as a gas detector, gas monitor, or gas sensor node, but the recommended terminology is gas detector.Gas Sensor CalibrationGas sensor calibration involves exposing a gas sensing element, module, or detector to various gas concentration standards under application-specific conditions. The output is then adjusted to match these standards, and the sensor is calibrated when the output aligns with the tested gas concentrations.Calibration provides information on uncertainty, non-linearity, and other parameters such as tested gases and their concentrations, and interferents such as ambient humidity and temperature and other gases. CoGDEM Guide to Gas Detection [3] describes common calibration routines and the factors affecting calibration. Companies that serve industrial safety markets also describe the necessary equipment and calibration technologies and offer guidelines and tutorials [4-6].A gas detector should operate under the expected application conditions to ensure it reports gas concentrations accurately. Examples of such conditions may be indoors, urban outdoors, industrial outdoors, exhaled breath analysis, and others. Accuracy refers to how closely the sensor readings match the actual gas concentrations. A calibration plan should account for all quantitative effects of these conditions. Figure 2 illustrates various gas detector accuracy levels and presents both complete coverage (full factorial) and partial coverage (fractional factorial) calibration plans [7]. These plans help manage calibration costs and complexity.Figure 2: Examples of full factorial and fractional factorial calibrations and correlation plot of benchmark versus sensor responsesThe calibration parameters are saved on the gas detector to convert measured responses into gas concentrations. In addition to traditional algorithms for industrial safety gas sensors, multivariate machine learning algorithms are also emerging. These algorithms can incorporate responses from the gas sensing element, along with contextual inputs from auxiliary sensors for interfering gases, temperature, humidity, and pressure. This data can be available on-board of the gas sensing module, on-board of the gas detector, or through the cloud. The calibration cost significantly increases the total production cost of mass-produced gas detectors because it adds additional steps and is time- and cost-intensive, but new approaches based by advanced algorithms have emerged that reduce calibration time and lower calibration costs [8, 9]. Not all gas detectors report individual gas concentrations. Some gas sensing applications only detect when a gas level reaches a specific threshold, for example a residential carbon monoxide alarm. Other gas detectors provide multiple air quality index values by accurately measuring concentrations of five pollutants: ozone, carbon monoxide, nitrogen dioxide, sulfur dioxide, and particulate matter of 2.5 mm and 10 mm [10]. These detectors are typically used for urban outdoor applications. Limitations of Single Output Gas Sensing Elements Prevent Their Acceptance in Many New AreasAvailable amperometric electrochemical sensors, semiconducting metal-oxide chemical resistors, pellistors, thermal conductivity sensors, and many others utilize gas sensing principles that were developed between the 1930s and 1970s. These innovations marked a significant advancement over using canary birds to detect carbon monoxide or miner’s lamps to identify methane in coal mines and other harsh environments. Early sensors relied on strong signals from a sensing element that measured high concentrations of a gas. Over time, engineers miniaturized these sensing elements without changing their underlying design principles, and today, the same sensing methods are still widely used to detect relatively high concentrations of specific gases for safety applications. However, single output gas sensing elements cannot mathematically differentiate various gases that produce similar sensor signals. They also cannot identify different sources affecting the sensor signal. Further, as the measured gas concentrations are getting smaller, the response drift of the sensing element and effects from other gases in air become more pronounced, reducing detector accuracy.For this reason, the United States Environmental Protection Agency recently highlighted that existing gas detectors have inherent limitations crucial to understand before collecting and interpreting data [11]. A Nature Perspective also noted that ambient interferences could make the data from single-output gas sensors “essentially meaningless” [12].Next Generation MEMS Gas Sensors: Solutions for Accurate and Stable Calibrated Gas Sensing Gas sensor developers and manufacturers are creating solutions that maintain accurate gas sensor performance at lower target gas concentrations over extended periods of time for chemically complex environments without increasing the hardware size or the amount of power consumption [13-15]. These solutions, inspired by traditional gas chromatography, mass spectrometry, and laser spectroscopy detectors that have exceptional gas-recognition abilities that are enabled by one or several independent response variables, which are factors that are varied in the detector in a controlled fashion. Examples of independent response variables are retention time in gas chromatography, mass-to-charge ratio in mass spectrometry, and wavelengths in laser spectroscopy. Although these traditional detectors are relatively large and expensive, they deliver significant societal benefits in exquisite multi-gas detection in chemically complex environments that have earned three Nobel Prizes [16]. SEMI’s MSIG Device Working Group is exploring the tremendous opportunity to emulate the mathematical principles of these large and expensive traditional gas detectors for miniature gas sensors [17, 18]. Our approach is to move beyond the limitations of single-output gas sensing elements and to preserve accurate gas detection in diverse operational scenarios. Right now, scientists and engineers are designing the next generation of miniature gas detectors to operate one or more gas sensing elements under measurement conditions that suppress or eliminate ambient interferences, boost stability, and reduce power consumption. To enhance gas sensing accuracy and stability, metal oxide gas sensing elements are modulated by gas sensing modules using temperature modulation [19] (e.g., Bosch, Renesas, 3S Technologies), dielectric excitation [20] (e.g. GE Vernova) or photoactivation [21] (e.g. N5 Sensors). Miniature electrochemical sensors use bias modulation and incorporate multi-frequency impedance enhanced readouts [15], whereas acoustic resonant sensing elements are modulated by temperature and multi-frequency impedance enhanced readouts of multiple harmonics [22]. Additionally, stable multi-element, multi-pixel, and multi-modal gas sensing elements are combining different sensing principles to gather more information from the same event, delivering more accurate responses. Figure 3 conceptually shows how these next-generation gas sensors should be able to compete with the traditional large and expensive analytical instruments on performance without the burden of high SWaP-C (size, weight, power, and cost) [23].Figure 3: Next generation gas sensors competing with exquisite performance of traditional analytical instruments, but without their high SWAP-C burden. SWaP-C stands for size, weight, power, and cost.ConclusionThe next generation of gas detectors promises to deliver high performance in diverse and complex environments by overcoming the limitations of single-output gas sensing elements and integrating advanced algorithms and multi-modal sensing techniques. These innovations will not only enhance safety and quality of environmental monitoring, but they will also open new applications for various industries. As the gas sensing field continues to evolve, interdisciplinary collaborations and continued research will be critical for unlocking the full potential of these technologies.Radislav A. Potyrailo is a Sr. Principal Scientist at GE Vernova Advanced Research.Andreas Schütze is a Professor at Saarland University.Sreeni Rao is a VP of Product Management and GM of Environmental Sensing at Interlink Electronics.Christian Meyer is a Sr. Manager of Application Engineering at Renesas Electronics Corporation.Paul Carey is Director of MEMS Sensors Industry Group at SEMI.References1. SEMI MS14 - Guide for Critical Parameters of Gas Sensors. SEMI: 2022; https://store-us.semi.org/products/ms01400-semi-ms14-guide-for-critical-parameters-of-gas-sensors.2. Rao, S.; Potyrailo, R.; Sakauchi, R.; Carey, P., SEMI MS14-0422 Standard: Critical Parameters of Gas Sensors For Emerging Applications. SEMI Advanced Sensors Seminar Series: 2023; p https://www.semi.org/sites/semi.org/files/2023-05/SEMI-Gas%20Std%20Webinar%20V14_230531.pdf.3. Greenham, L., The CoGDEM Guide to Gas Detection. ILM Publications: 2012.4. Gas Detector Calibration Procedures, Requirements and Tips, Industrial Scientific 2025, https://www.indsci.com/en/blog/gas-detector-calibration.5. Gas Detector Bump Test: Bump Testing and Calibration of your Gas Monitors, PK Safety 2025, https://pksafety.com/blogs/pk-safety-blog/bump-testing-and-calibration-of-your-gas-monitors.6. What Are Calibration and Bump Tests for Portable Gas Detectors: Key Differences and Ways to Help Streamline Compliance, MSA 2024, https://blog.msasafety.com/what-are-calibration-and-bump-tests-for-portable-gas-detectors/.7. Ryan, T. P., Modern Experimental Design. Wiley: Hoboken, NJ, 2007.8. Fonollosa, J.; Fernandez, L.; Gutiérrez-Gálvez, A.; Huerta, R.; Marco, S. Calibration transfer and drift counteraction in chemical sensor arrays using direct standardization, Sens. Actuators B 2016, 236, 1044-1053.9. Robin, Y.; Amann, J.; Schneider, T.; Schütze, A.; Bur, C. Comparison of Transfer Learning and Established Calibration Transfer Methods for Metal Oxide Semiconductor Gas Sensors, Atmosphere 2023, 14, (7), 1123.10. Technical Assistance Document for the Reporting of Daily Air Quality – the Air Quality Index (AQI), US Environmental Protection Agency 2014, EPA-454/B-24-002.11. Barkjohn, K. K.; Clements, A.; Mocka, C.; Barrette, C.; Bittner, A.; Champion, W.; Gantt, B.; Good, E.; Holder, A.; Hillis, B. Air Quality Sensor Experts Convene: Current Quality Assurance Considerations for Credible Data, ACS ES T Air 2024, 1, (10), 1203–1214.12. Austen, K. Pollution patrol, Nature 2015, 517, 136-138.13. Bur, C.; Bastuck, M.; Spetz, A. L.; Andersson, M.; Schütze, A. Selectivity enhancement of SiC-FET gas sensors by combining temperature and gate bias cycled operation using multivariate statistics, Sens. Actuators, B 2014, 193, 931-940.14. Schütze, A., Keynote: High performance gas measurement systems – bridging the gap between sensors and analytics. IEEE International Symposium on Olfaction and Electronic Nose (ISOEN), Grapevine, TX, May 12-15: 2024.15. Potyrailo, R. A. In Cross-Pollination of Electronics and Mathematics: Unlocking New Horizons in Ambient Gas Sensing, SEMI MEMS and Sensors Technical Congress (MSTC) 2025, Georgia Institute of Technology, Atlanta, GA, March 26-27, 2025.16. The Nobel Foundation 2025, https://www.nobelprize.org/prizes/lists/all-nobel-prizes/.17. Potyrailo, R. A.; St-Pierre, R.; Crowder, J.; Scherer, B.; Cheng, B.; Nayeri, M.; Shan, S.; Brewer, J.; Ruffalo, R. First-order individual gas sensors as next generation reliable analytical instruments, Appl. Spectrosc. 2023, 77, (8), 860–872.18. Potyrailo, R. A.; Shan, S.; Cheng, B. Individual Optical Multi-Gas Sensors as Next Generation Second-Order Unobtrusive and Continuous Operation Analytical Instruments, Microchim. Acta 2025, Special Issue in Memory of Otto S. Wolfbeis, DOI: https://doi.org/10.21203/rs.3.rs-6234291/v1.19. Schütze, A.; Sauerwald, T., Dynamic operation of semiconductor sensors. In Semiconductor Gas Sensors, Elsevier: 2020; pp 385-412.20. Potyrailo, R. A.; Go, S.; Sexton, D.; Li, X.; Alkadi, N.; Kolmakov, A.; Amm, B.; St-Pierre, R.; Scherer, B.; Nayeri, M.; Wu, G.; Collazo-Davila, C.; Forman, D.; Calvert, C.; Mack, C.; Mcconnell, P. Extraordinary performance of semiconducting metal oxide gas sensors using dielectric excitation, Nat. Electron. 2020, 3, 280–289.21. Deb, S.; Mondal, A.; Reddy, Y. A. K. Review on development of metal-oxide and 2-D material based gas sensors under light-activation, Current Opinion in Solid State and Materials Science 2024, 30, 101160.22. Potyrailo, R. A., Tutorial: Next generation of gas sensors: anticipated and unanticipated advantages over last-century sensor designs. IEEE SENSORS, Vienna, Austria, Oct 29 - Nov 01: 2023.23. What is SWaP-C?, NSTXL National Security Technology Accelerator 2022, EPA-454/B-24-002.

Dr. Sondra Hellstrom of Bosch speaks to Tim Brosnihan of SEMI MEMS & Sensors Industry Group (MSIG) to preview her presentation on decarbonization at MEMS & Sensors Technical Congress (MSTC), May 23-24 at MIT in Cambridge Massachusetts. 

MSTC 2023 will open with Nora Houlihan, senior research analyst for MEMS & Sensors with Omdia, providing a 2023 market update and a look at what’s ahead.

Companies of all sizes—not just industry juggernauts—are vital to the long-term health and well-being of that ecosystem. We hope you’ll stake your claim when the time comes. We all stand to benefit if you do.

More consumer and industrial products may emit volatiles that now are known to be harmful, including furniture, passenger cars, and industrial trucks. Interest in detecting and measuring the gaseous pollutants is growing in order to create effective responses to reduce or eliminate health risks. 

The latest MEMS and sensors advancements in markets ranging from displays to biotech and in key areas of manufacturing including sensor devices and fabrication came into sharp focus at the 2022 MEMS & Sensors Technical Congress (MSTC) in late April as 120+ industry experts gathered in Berkeley, CA.

To help build the MEMS and sensors industry’s understanding of the Smart City challenges facing city technical and management personnel (and their integrators), MSIG held an invitation-only Sensorize Your City workshop for sensor device manufacturers and integrators on April 5 in Columbus OH.

What is analog computing and what do engineers need to know about it so they can make sound decisions during product development to differentiate their battery-operated always-on devices?

Areas packed with dense foliage. Mile-deep mines and tunnels. Urban canyons. Indoor environments. Global Positioning System (GPS) technology has long been a boon to location tracking of aerial, terrestrial and aquatic vehicles — as well as to people in motion — but in many cases it can’t function with a high degree of reliability, either because the GPS signal is somehow obstructed, or worse, is jammed or spoofed. Delivering higher precision and higher reliability in GPS-denied environments — as well as immunity to jamming and spoofing — positioning, navigation and timing (PNT) represents the next evolutionary step in location positioning and tracking. With PNT so critical to a range of defense, commercial and industrial applications — and with sensors the building blocks of PNT solutions —the MEMS Sensors Industry Group, a SEMI Technology Community, is ensuring that our members play a transformative role in PNT innovations. We’ve secured $14.9 million in research dollars for PNT R D over the past 18 months, marking Phase I of a project funded through a public-private consortium with the U.S. Army Research Laboratory (ARL). With the typical funding structured as a 50/50 cost share with the industry participant, the research dollars go farther, and the level of commitment that each recipient makes is more pronounced. As we look ahead to Phase II of the MSIG PNT R D project, the details of which we’ll announce later this year, we’d like to reflect on the companies and research labs that won bids through a competitive process supported by the SEMI-MSIG PNT Technical Advisory Council and the SEMI-MSIG PNT Governing Council. Winners submitted proposals that both met our criteria for advancing PNT technologies relative to mobility, size and weight, and that laid a path toward greater cost efficiency and lower product price. “PNT doesn’t displace GPS,” said Tim Brosnihan, executive director of SEMI-MSIG. “Rather, getting the two technologies to work together improves position and tracking. While current PNT solutions use inertial measurement units, or IMUs, to effectively maintain positioning accuracy in the absence of a GPS signal, it’s also true that accumulated bias and noise-related errors in the IMUs make positional determination unreliable. Like most great pairings, GPS and PNT can work together. We can use IMUs when GPS is unavailable, and when GPS returns, it can be used to reset the IMU errors. So when the GPS signal is lost again, the IMU can maintain navigation and location. “We’re focusing this PNT project on technologies that will allow accurate positional determination in the absence of a reliable GPS signal for prolonged periods,” added Brosnihan. Here are snapshots of the 10 companies and research institutions that won awards for their PNT-focused developments. Analog Devices is developing an optimal size, weight, power, and cost (SWaP-C) solution for applications requiring high-accuracy navigation and uncompromised reliability. The company’s mode-matched navigation-grade gyroscope with system ID leverages an innovative sensor and its associated process design, a robust high-volume manufacturing flow, and system-control algorithms to achieve very high-performance (0.01 degree/hour bias instability and 0.005 degree/√hr angle random walk). Carnegie Mellon University (CMU) is developing a CMOS MEMS high-stability accelerometer through machine learning (ML). If embedded in footwear, these ML-optimized accelerometers could be used in personal navigation. If embedded in a golf ball, baseball or hockey puck, the accelerometer could extract the trajectory of the object in motion by measuring its shock (force). The CMU device validates state-of-the-art performance of the university’s high-dynamic-range accelerometer systems-on-chip. It also validates and tests ML models by measuring the accelerometer and auxiliary sensor output over long time periods (e.g., 1 hour, 10 hours, days) to collect independent long-duration time-series data. By modeling drift from environmental influences — along with possible overall system changes from extreme events, such as high-temperature excursions and shock — designers can dramatically reduce navigation errors to support more accurate navigation over longer time periods. GE Research is developing a novel MEMS gyrocompass that will enable high-end north-finding systems, traditionally unaffordable for automotive and consumer applications. The device will be available in mass-market applications such as robotics and autonomous vehicle navigation in GPS-denied environments. The MEMS gyrocompass enables a 10x reduction in SWAP-C with high accuracy. An additional benefit of this work is that GE will offer a foundry service process development kit (PDK) for its Polaris MEMS process, speeding the development and manufacture of MEMS devices in an advanced processing facility. Georgia Institute of Technology is developing high-aspect-ratio monocrystalline silicon carbide-on-Insulator (SiCOI) MEMS devices that will reduce navigation angle errors, potentially making widescale pedestrian navigation available in mass-market applications such as smartwatches and smartphones. The platform for ultra-high-performance bulk acoustic wave (BAW) gyroscopes and timing resonators will feature material properties that allow a much better structural symmetry and a higher-resonant quality factor (Q) than silicon MEMS (Si MEMS). Honeywell is working to enhance the navigation accuracy of commercial and military vehicles in GPS-denied environments through an innovation that dramatically improves the performance of a MEMS IMU by both refining candidate ML algorithms, including recurrent neural networks (RNNs), and by combining deep neural network (DNN)-based calibration and sensor fusion algorithms. PARC is developing a new materials platform for photonic integrated circuits (PIC). Aluminum gallium nitride (AlGaN), an ultra-wide bandgap semiconductor, is epitaxially grown to produce single-crystal layers for fabrication of optical components, such as waveguides and micro-ring resonators for optical signal processing. The project includes design and fabrication of specialized laser diodes at wavelengths needed to probe qubits based on atomic ions (e.g., strontium and ytterbium). The new platform offers several benefits: low optical loss from the ultraviolet (UV) to infrared spectral bands excellent non-linear optical properties for efficient frequency-generation processes (e.g., optical frequency combs); and enabling technology to realize compact, field-deployable quantum systems for PNT applications, such as ultra-fast distance measurements, microcombs for optical atomic clocks, photonic radar, optical coherence tomography, and coherent communications — all applications that benefit from the lower cost and small chip size of these integrated photonic circuits By expanding its proprietary EpiSeal encapsulation process to include new materials and topologies, SiTime is developing low-impedance and low-noise MEMS resonators with an ultra-stable wafer-level package. Because these novel MEMS resonators are highly reliable and very compact — while using less power and providing lower RF noise — they’re ideal for 5G RF timing applications, IoT devices, and smart vehicles. Teledyne Scientific Imaging (CSAC project) is conducting a study to identify paths to reduce the cost of battery-operated chip scale atomic clocks (CSAC) that provide affordable precision timing for denied environments. The project goal is to identify viable paths of reducing cost by an order of magnitude, without sacrificing performance. In addition to exploring design and manufacturability solutions, project researchers are performing short loop experiments as proof-of-concept validation. Through a second award, Teledyne Scientific (IMU project) is advancing packaging and integration for compact, navigation-grade six degrees of freedom (DOF) MEMS IMUs. Featuring reduced bias instabilities associated with packaging stresses and ambient temperature influences, the Teledyne Scientific IMUs promote environmentally robust low-stress packaging of wafer-level vacuum packaged (WLVP) MEMS gyro resonators, facilitating a lower-cost, smaller and more accurate IMU for performance-driven PNT applications. Twinleaf is developing a new light source module ideally suited for integration directly into quantum sensors. This project integrates a bright, tunable distributed Bragg reflector (DBR) near infrared (IR) 795nm wavelength laser made by the project’s subcontractor (Photodigm) into a package that locks the laser to an atomic reference line in a microfabricated vapor cell. The laser module’s high-output intensity and low magnetic signature will enable breakthrough performance levels for Twinleaf’s magnetometer and other quantum sensors requiring the light source integrated into the sensor module. Request for Proposal for Phase II of SEMI-MSIG PNT Program Opens Q4 2021 SEMI-MSIG will accept request for proposal (RFP) submissions for Phase II of its PNT program starting in Q4 2021. This year, in addition to funding IMU and timing device projects, MSIG will also consider proposals on imaging-based navigation solutions. If you’d like to submit for Phase II, sign up to receive more information on the RFP by visiting SEMI’s R D Programs page. You can also connect with Paul Carey by email, [email protected] or LinkedIn. Paul Carey, Ph.D., is the director of the MEMS Sensors Industry Group. With deep domain expertise in X-ray imaging backplane platforms — and their supply-chain technologies such as flexible substrates, laser annealing for semiconductors and silicides, thin film transistors (TFT) for flexible OLED displays, and polysilicon-on-plastic TFT technology — Carey has held technical leadership positions at dpiX, Applied Materials, and Lawrence Livermore National Laboratory. He received a double-major B.S. from UC Berkeley in Electronical Engineering and Computer Science (EECS), and Materials Science and Engineering (MSE). Carey holds an M.S. in EECS from UC Berkeley and a Ph.D. in MSE from Stanford University.

Our home state of California has wilderness areas of extreme climates, from desert to high-altitude snow-capped mountains. Every year, people need rescue because they’ve ventured into the wilderness without proper training, or even essential gear, such as water or a warm parka. Many MEMS product development teams get a similar start. They undertake a challenging multi-year journey without enough of the most precious resource needed for success – enough money to finish. As a MEMS product development firm that's completed more than 400 projects, 25% of these with startup companies, we’ve been having many of the same conversations about the road to commercialization with MEMS entrepreneurs. Through that experience, we’ve seen a share of these entrepreneurs – many of whom come right from graduate programs or from outside the semiconductor industry – experience disappointing outcomes. And it’s not because of their technology. Rather, their lack of familiarity with electronic product integration and wafer-based manufacturing often influences a too-optimistic development plan that doesn’t factor in enough time and budget. As technologists, it was tough for us to see companies with promising young technologies struggle due to lack of planning or funding. We would like to see more entrepreneurs succeed, and that was the impetus for our new book, MEMS Product Development: From Concept to Commercialization, some highlights of which follow. Time and money What can MEMS startups do to pave the way for a successful commercial launch, particularly when a long period of scaling up manufacturing is often needed during the go-to-market process? Since money is usually a limited resource, it’s important to prepare a realistic development timeline supported by sufficient funds allocation from the start. As this is easier said than done, we’ve seen both startups and established companies make common financial blunders during MEMS product development. These include: Reserving inadequate funding for developing the entire MEMS product, including packaging, electronics and software Creating an unrealistic timeline for development, resulting in a cash-flow problem Only securing enough funding for the first run at a foundry when, in fact, numerous runs are far more typical Unplanned gaps of months or more between funding tranches, which slows momentum Based on our varied client experience, the 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) requires on average four years and US$4 million. And that’s just for engineering. Business administration, sales, marketing and other company costs are additional. While it’s common to spend much more for more complex devices or product systems, it’s rare to spend less, unless you’re working with existing IP, such as a foundry process platform, which also could accelerate development time. Typical engineering-only budget required to develop a MEMS product through four stages of development, to the point of volume-production readiness. Reprinted with permission from MEMS Product Development: From Concept to Commercialization (Springer, 2021). Don’t go thirsty in the desert No one wants to get stranded in the desert without enough water, which is why it’s so important to carefully articulate your timeline and secure adequate funding before starting MEMS development. In MEMS development, just as in wilderness adventuring, things rarely go exactly as planned: Wafers break, engineers take a long time to debug, customers change their minds, and random events like storms (or a pandemic) disrupt supply chains. That’s why adding some buffer to your development timeline and your budget will sustain your company through the inevitable delays and setbacks. Plus, there’s generally a ripple effect to a delayed new-product introduction. A slower-than-predicted launch places a burden on a company’s finances because the fixed overhead costs of the entire organization will continue to consume cash while waiting for product launch. Not a lump-sum game Although you might wish to receive one big funding check when you get started, the reality is that investors and executives won’t provide the entire development funding in one sum. They give money in tranches, generally demanding you meet some pre-determined criteria or demonstrate set benchmarks before they’ll release more funds. To best manage tranche funding, a company must carefully plan and set their investors’ expectations for realistic outcomes in advance. A common crisis for startup companies occurs when investors only provide enough budget to execute the very first run at a foundry and then demand to see functional chips before providing the next tranche. However, the aim of the foundry’s first wafer run isn’t to produce working chips. It’s to begin the year-long process of setting up for high-volume manufacturing. The first wafers are unlikely to yield well, or at all, putting the startup at great risk with its frustrated investors. Setting investors’ and executives’ expectations correctly from the start, realistic budgeting and having regular communication about progress and upcoming needs all help to keep the money flowing. Any gaps in funding will waste valuable momentum, which ultimately leads to more expense and delay in the overall product development. It can become especially damaging when the wait for money forces the foundry to stop work, because processes go stale after a few months, and also during busy times, when your product could be sent to the back of the foundry’s queue. As experienced outdoors people know, to enjoy wilderness adventures, you need to plan where you’re going, anticipate common risks, and then prepare accordingly. It’s the same in MEMS product development. Having a good grasp of your timeline and realistic expectations about the funding required to reach commercialization are essential steps in a successful journey. Want to learn more about MEMS Product Development: From Concept to Commercialization (Springer, 2021)? Order the book based on A.M. Fitzgerald Associates’ extensive experience helping entrepreneurs and other innovators commercialize their MEMS devices. Alissa M. Fitzgerald, Ph.D., founded A.M. Fitzgerald Associates, LLC, a MEMS product development firm based in the Bay Area, California, in 2003. She has over 25 years of engineering experience in MEMS design and fabrication and now advises clients on the entire cycle of MEMS product development, from business and IP strategy to supply chain and manufacturing operations. Carolyn D. White, Ph.D., has a background in mechanics of materials and specializes in the design and fabrication of MEMS devices for a wide range of applications. She has additional experience in foundry transfers and technology strategic analysis, including of the evaluation of patent portfolios, feasibility studies, and cost/performance analysis. A.M. Fitzgerald Associates (“AMFitzgerald”) is longtime member of MEMS Sensors Industry Group®(MSIG), a SEMI technology community that connects the MEMS and sensors supply network in established and emerging markets to enable members to grow and prosper. Visit us today.