Who’s Driving the MEMS Evolution Revolution Now? (Part 2 of 3)

I am pleased to bring you the second part of a three part series on the MEMS Evolution Revolution, written by my colleague, and long-time MEMS industry insider, Howard Wisniowski.  So far in this series, Howard has taken us with him to “visit” member company Qualtré, and taught us about bulk acoustic wave (BAW) solid state MEMS gyroscopes.  In part 2, we will begin to learn about radio frequency (RF) MEMS, an innovative application called “Tunable Antennae”, and a start up who is pioneering the advances of this new technology.

I hope you are as excited as I am to read this series and I welcome you share your stories of other MEMS start ups that are breaking out in their own markets, whether it be in agriculture or acoustics; healthcare or helicopters. MEMS truly is everywhere and it’s likely the innovative smaller companies who will spread it further, faster and for longer. Viva la Revolution!

Who’s Driving the MEMS Evolution Revolution Now?

Part 2 of 3

Howard Wisniowski, Freelance Editor

What’s most exciting about MEMS technology is watching how it is evolving. As a participant in the MEMS industry for over 15 years, I have witnessed much of the evolution and revolution take place. In Part 1, I highlighted an innovative and disruptive inertial MEMS technology referred to as bulk acoustic wave (BAW) technology. This new class of solid state stationary gyroscopes is opening up many new application possibilities by being able to meet the performance, size, cost, and reliability requirements for many emerging MEMS inertial sensor applications.

Part 2 focuses on radio frequency (RF) MEMS and a very innovative and disruptive application referred to as tunable antennae. It is hard to believe that one of the most important parts of a mobile phone is the antennae, which is very low-tech. With today’s smartphones that incorporate very sophisticated technology from gazillion-transistor CPUs controlling everything to state-of-the-art retina display on the front ends, the antennae for GSM, LTE, WiFi, and Bluetooth, are simply pieces of metal.

We all can recall when devout iPhone followers were outraged by the fact that an Apple device could be defeated when water-filled, fleshy fingers touched the metal antenna, it attenuated (weakened) the signal and resulted in dropped calls. The fact of the matter is that every smartphone has similar issues. Fortunately, for every mobile device maker, there’s an alternative to normal antennae: RF MEMS.

RF MEMS, as the name suggests, are semiconductor chips that can alter their physical (mechanical) state with the application of movable structures. When applied to an antenna, RF MEMS can be used to make antennae that automatically tune and re-tune themselves to both incoming and outgoing signals. For example, if one should put a finger on an RF MEMS antenna it can automatically re-tune itself so that no calls are dropped. What’s more, this is an emerging application where IHS iSuppli has reported that sales of RF MEMS devices are could reach $150 million by 2015.

RF MEMS Antenna Tuners

At WiSpry, a start up in Irvine, CA and another MIG member, they are pioneering advances in the field of tunable RF technology and addressing the emerging needs of modern smartphones.  Today’s smartphones have a number of radios to deal with — GSM, 3G, CDMA, W-CDMA, LTE, Bluetooth, WiFi, and even FM and TV radios in some cases. Each one has its own silicon circuitry and usually its own antenna too. Additionally, there are now a burgeoning number of frequency bands needing to be supported for 4G LTE cellular – ranging today from 700 Mhz to around 3700 Mhz. What’s more, the 3GPP standards are now allowing more than 43 different frequencies and there is an emerging demand for “Carrier Aggregation” in LTE – Advanced, the newest set of standards, which will have simultaneous “aggregation” of multiple frequencies on a single phone, allowing huge bandwidth improvements.

WiSpry’s RF MEMS-based antenna tuner technology will play pivotal roles in these advancements by potentially enabling devices with just a single antenna and transceiver. By reducing the number of necessary components in a handset while allowing the radio front-end to be programmed to work in any frequency band and with any radio standard using the same set of hardware, a “World-Phone” architecture is possible and truly disruptive. Finally thanks to MEMS, the antennae on mobile devices will actually function more efficiently as they were initially intended – to carry and convey data and yes, even your phone calls.

What is Healthcare without MEMS?

I might have a problem.  My heart rate is 150 BPM, which means I might be suffering from tachycardia according to WebMD.  Or maybe I’m just exercising and am well within my heart rate range.  Without the context, that 150 BPM number is absolutely worthless.

To understand what the number really means, a doctor is going to have to rely on the information I tell her.  And let’s be honest; I’m a liar.  We all are when it comes to our physical fitness.  We all omit that chocolate cake we had for breakfast while exaggerating that nightly run around the block into a full mile.  It’s called the Hawthorne effect, and it means we’re generally useless when it comes to providing context for any of our biological metrics.

That’s where MEMS comes in.  In general, MEMS sensors track motion within a 3D space.  The figure below shows a 10-axis frame of reference, which means that by using four different MEMS devices (Accelerometer, Gyroscope, Magnetometer, and a Pressure Sensor), this device could potentially track motion along the XYZ axis along with the up and down pressure sensor.

Freescale
Source: Freescale

This ability to track motion is revolutionizing our healthcare.  We don’t even need to add all these sensors, though each one does provide an extra layer of context.  By combining something as simple as an accelerometer (the average cost for a MEMS sensor is between $1 – $2) with a heart rate monitor, we automatically know the context for someone’s heart rate and can determine if that 150 BPM mentioned earlier is spiking when we’re sleeping or if we’re in the middle of a five-mile run.

In order to make a system like this work, we are also going to need a low power MCU to aggregate the data the sensors are collecting.  Right now the most popular part is a 16bit MCU that costs about $0.60.  Manufacturers are liking this part because of its low cost and low power consumption, while having just enough memory to not be overkill.  These wearable appcessories are then going to send this aggregated data to a tablet or smart phone, where a more complex sensor fusion could take place.  TI has a popular low cost Bluetooth chip (CC2540) that is under $2, so all together, this is a very low cost system that aggregates a ton of personal data and transmits it to a dashboard for our consumption (smartphone or tablet).

Texas Instruments is one of the vendors behind these types of systems.  They have relationships with several MEMS vendors in order to off their SensorTag development kit and reference design for Bluetooth low energy sensor applications.  The figure below uses TI’s CC2541, a 2.4GHz Bluetooth low-energy SoC combined with the Sensirion SHT21 humidity sensor, InvenSense’s IMU3000 Gyroscope, the Kionix KTXJ9 accelerometer, Freescale’s MAG3110 magnetometer, and the Epcos TS400 barometric pressure sensor.  A product like this would take in the aggregated data from a healthcare appcessory and create a context-rich dashboard for the consumer or doctor.

Texas Instruments
Source: Texas Instruments

This is how easily MEMS is changing people’s lives for the better.  On a larger scale, MEMS has the potential to change entire societies.  Through fitness and health apps, the industry can aggregate everyone’s personal health stats and translate those into actionable items for communities.  Imagine a community where everyone is wearing a Fitbit and the entire town knows who has the best walking rates.  What if neighborhoods had their own fitness competitions?  Towns can take these stats and compare them with national rates of obesity or illness and decide to take more relevant action like creating more walking and biking paths.  In this manner local communities could choose to divert funds to the areas where the data tells them they are falling behind the curve.

When we talk about Big Data, this is what we’re talking about, and none of it is possible without the context MEMS provides.  To find out just how large this market is going to be, read Semico’s recently completed Mobile Healthcare Study (hint: over 270 million units by 2017).

Improved electronic compass software released: Xtrinsic eCompass software

Originally posted by Michael E Stanley of Freescale Semiconductor in The Embedded Beat on April 29, 2013

A few weeks ago, my coworkers and I had the pleasure of participating in an awards ceremony in which Electronic Products Magazine presented Freescale with a Product of the Year award for our Xtrinsic eCompass software. This software processes the outputs of two sensors (an accelerometer and a magnetometer) to implement a tilt-compensated electronic compass. The software is available in source code format supported by an easy-to-use click through license.

We were at revision 2.0 of that library when Electronic Products announced the award. Since then, revision 3.0 has been uploaded to our web site. To download, click on the ECOMPASS_SW link on the Xtrinsic eCompass software page, read and approve the license agreement that pops up (you can freely use this software in products which include Freescale magnetometers and accelerometers), and save the offered .zip file onto your hard drive.

 

When you expand the zip file, you will have a folder called “eCompass”, with sub-folders: “Documents” and “Software”.  This is a major new release. I’m going to claim it to be the best-documented e-compass solution anywhere, thanks to the Herculean efforts of my good friend Mark Pedley, who also supplied much of the content for this post. Here’s what you’re going to find in the “Documents” folder:

  1. Software for Tilt-Compensated eCompass with Magnetic Calibration (v3 Release) User Guide
  2. AN4676 – Euler Angle, Rotation Matrix and Quaternion Representations of Orientation in Aerospace, Android® and Windows 8® Coordinates
  3. AN4684 – Magnetic Calibration of Hard and Soft Iron Interference
  4. AN4685 – Tilt-Compensated eCompass in Aerospace, Android and Windows 8 Coordinate Systems
  5. AN4696 – Accelerometer and Magnetometer Sensor Simulatoin for Tilt-Compensated eCompass
  6. AN4697 – Low Pass Filtering of Orientation Estimates
  7. AN4698 – CPU, Flash and RAM Benchmarks : Xtrinsic eCompass and Magnetic Calibration Algorithms
  8. AN4699 – Data Structures for Matrix and Vector Algebra
  9. AN4700 – Control Loop, Data Structures and Compile Time Constants
  10. AN4706 – Accelerometer and Magnetometer Selection and Configuration

 

The software itself has been expanded and improved.

Flow chart.jpg

The “Software” directory contains half a dozen C source and header files that contain everything you need to implement your own e-compass. It also includes a pre-compiled command line tool that lets you simulate performance of the e-compass. The ANSI C source code is processor agnostic allowing Freescale customers to retain their existing MCU architecture. The software is highly optimized to minimize use of program memory, RAM and floating point calculations. Software features include:

 

  • The code compiles into 10KB of ARM Thumb2 object code and uses less than 4KB of RAM.
  • A dedicated floating point unit (FPU) is not required and the software can run on typical 32 bit integer processors with software floating point emulation.
  • Orientation is provided in Euler angle (roll, pitch, yaw and compass heading), rotation matrix and quaternion formats.
  • Supports Aerospace, Android and Windows 8 coordinate systems
  • Tilt-compensated
  • Programmable low pass filter
  • Quality of fit metric indicates expected compass heading error
  • Resilient to magnetic jamming corrupting calibration
  • Three levels of hard and soft iron magnetic calibration are provided at increasing levels of performance and computational complexity.
    1. The simplest 4 element calibration solver computes the hard iron correction vector and geomagnetic field strength and removes the largest component of the magnetic interference caused by ferromagnetic components on the circuit board. It consumes 3300 floating point operations per call.
    2. The seven element calibration solver corrects for differing magnetic permeability along the three Cartesian axes and is suitable for the more complex calibration environments found in the dense circuit board layouts of smartphones and tablets. It consumes 20,000 floating point operations per call.
    3. The 10 element calibration solver computes a best-fit solution to the 10 dimensional magnetic optimization problem including off-diagonal elements of the soft iron matrix. It consumes 62,000 floating point operations per call.

The web-release includes source for options 1 and 2 above. Option 3, the highest performing 10 element calibration solver, is not available in source form, but is available under license in object code format for ARM Thumb2 processors.

 

If you have used previous generations of our e-compass software, you will see major improvements in the feature set above. I like the fact that it now supports any of three different orientation representations right out of the box. The math behind an electronic compass isn’t easy, but Mark has done an excellent job of breaking it down into manageable chunks that are easily digested. So please, download the new release, give it a go and let us have your feedback.

 

References are available on the original blog post. 

Who’s Driving the MEMS Evolution Revolution? (Part 1 of 3)

I am pleased to bring you part one of a three part series on the MEMS Evolution Revolution, written by my colleague, and long-time MEMS industry insider, Howard Wisniowski. Howard takes us with him to “visit” three exciting MEMS startups that are breaking new ground in the mobile/consumer market. In part one, we learn about bulk acoustic wave (BAW) solid state MEMS gyroscopes and meet MIG member company Qualtré. In parts two and three we journey to find out what companies are driving the MEMS evolution revolution with their exciting nascent disruptive technologies. I hope you are as excited as I am to read this series and I welcome you share your stories of other MEMS startups that are breaking out in their own markets, whether it be in agriculture or acoustics; healthcare or helicopters. MEMS truly is everywhere and it’s likely the innovative smaller companies who will spread it further, faster and for longer. Viva la Revolution!

 

Who’s Driving the MEMS Evolution Revolution Now?

Part 1

Howard Wisniowski, Freelance Editor

Like the transistor and the microprocessor, MEMS are often described as a disruptive technology, as in change-the-world, turn-it-upside-down, rewrite-the-rules-of-the-game. You can forget about this kind of incremental change, however, fitting easily into corporate business plans. Few, if any, roadmap processes are available to accommodate new innovative disruptive technologies that either have the potential to radically change the way products are currently being produced or are the foundation for products that might create entirely new industries, nascent disruptive technologies. Within many established corporate environments, roadmaps all too often focus on sustaining existing technologies with a mature sales base and use variations of tried and true processes that exist in their fabs. Start-ups don’t have these types of investments enabling them to build on the shoulders of their predecessors and develop products that take a fresh look at what benefits product design engineers are seeking for new and existing end applications.

Today on the “revolution” side, the demand for MEMS technology is still booming thanks to not only to the continued growth of high volume automotive and consumer applications where MEMS sensors have become mainstream, but also to the continued development of emerging applications in robotics, energy harvesting, and healthcare. On the “evolution” side, however, there are even more exciting and disruptive things going on with MEMS technology that is poised to drive the next wave of MEMS enabled products and applications. There are hundreds of companies, universities, and thousands of researchers around the globe working on MEMS projects. Many have the underlying technology that is well beyond the laboratory, ready for deployment, and are now seeking funding.

Highlighting this very active sector, Yole Development reports on the continuing growth of emerging MEMS products and applications. Alongside many of the old timers, their reports cite as many as 50 startups designing emerging MEMS devices that have the possibility to ramp up to large volumes quickly with growing access to contract foundries.

Within this large field, several new “disruptive” MEMS devices will be highlighted in this three part series beginning with bulk acoustic wave (BAW) MEMS technology. This new and disruptive MEMS technology is now being applied to innovative MEMS gyroscopes.

 

Bulk acoustic wave (BAW) solid state MEMS gyroscopes

According to analysts at IHS iSuppli, the MEMS gyroscope market displaced accelerometers as the revenue champion in consumer and mobile MEMS applications when revenue grew 66 percent from $394 million in 2010 to $655 million in 2011. While engineers now design systems that include MEMS gyros as essential components, particularly designers of mobile devices, suppliers are scrambling to meet their needs for low power, small size and low cost.

Qualtré, Inc. (Marlborough, MA) is one MEMS start-up and MIG member that is addressing these issues with an innovative MEMS technology referred to as bulk acoustic wave (BAW) technology. BAW technology is now being used to pioneer a new class of solid state stationary gyroscopes that not only meet power, size and cost requirements, but also add high performance to the mix. Unlike older MEMS gyro technologies that use moving masses vibrating at low frequency range of 5 to 50 kHz (I don’t want to get too technical here), BAW MEMS gyros operate in the megahertz frequency range (1‐10MHz), several orders of magnitude higher. This is enabled by the very stiff nature of the BAW technology. This stiffness not only results in MEMS gyros that are insensitive to vibration in the environment but also prevents stiction both in manufacturing and during operation in the field, thus removing a major yield and reliability problem found with the vast majority of other MEMS devices. These features results in improved performance in real world applications where vibrations are present and degrade the operation of current gyros.

By combining these performance advantages of the BAW sensor design and the scalability of Qualtré’s proprietary HARPSS™ process (High Aspect-­Ratio Combined Poly and Single-­Crystal Silicon), BAW MEMS gyros have also demonstrated very stable signals (aka low drift) which is important for pedestrian navigation, improved noise density for better resolution and more accurate measurements, and a wider dynamic range that expands detectable signals. This kind of innovation is what will drive the next wave of end-product product designs for new and existing applications.

Karen’s blog from MEMS Executive Congress: Part 2

I last left you hanging, waiting to hear more about the heated conversations between the panelists and the audience – and I have to tell you, it really started heating up in the audience during the energy panel. Ooo baby it was jumping.

MEMS Executive Congress Europe 2013MEMS in energy can mean a lot of things – and our panelists diverse perspectives discussed a great deal, but the majority of the audience wanted to focus on the topic of MEMS in energy harvesting. Though not necessarily experts in this field, thankfully our panelists were up to the challenge. Our moderator was Bert Gyselinckx, General Manager, Holst Centre, imec; Wim C. Sinke, Program Development Manager, Solar Energy, Energy Research Centre of the Netherlands; Eric Yeatman, Professor of Microengineering, Deputy Head of Department, Imperial College London; and Harry Zervos, Senior Technology Analyst, IDTechEx. I actually should probably add Rob Andosca of MicroGen Systems as a fifth panelist as he was eager to ask and answer any question from the audience with his BOLT energy harvester in hand.

I loved the diversity of perspective on this panel –Wim for example does not have an entirely MEMS-centric background. His expertise is in solar and photovoltaic energy and he spoke of how multiple technologies will work together to make reliable and sustainable energy system, as well as the importance of portfolio management – combining different energies in an active way to make it work. We in MEMS could learn a lot from guys like Wim (I hope everyone picked up his business card; I know I did).

The panel also spoke about wireless sensor networks and Harry gave a great overview of the three technologies that are converging: 1. Microgenerators and energy storage (vibration, solar, heat, tree resin, etc.); 2. Ultra low-power electronics (currently being developed) – helping power sensors; and 3. Transmission protocols that don’t need a lot of power to send data. Eric followed up with the poignant view that until things become truly wireless, you can’t really have wireless sensor networks. And once they are wireless how will they be powered – by energy harvesting or battery? This opened the floodgates and I, with microphone in hand had to jog all over the audience to capture the comments and follow-up questions from the audience.

Let me be diplomatic and say that there is no clear consensus out there on MEMS energy harvesting. And out came the very clever quotes including some of my favorites including this one from Wim: “Don’t look at MEMS as the energy harvesters, MEMS are the enablers to help realize energy savings.” And this one from someone (maybe you’ll remember and leave a comment here)  “I’m happy to hear everyone in MEMS talking about energy, but I can assure you that not everyone in energy is talking about MEMS…yet.” And Bert’s: “MEMS will probably not be main source of energy replacing nuclear power plants soon; but MEMS will enable increased intelligence in energy applications.” As great as these sound bytes were, the show stealer came when Rob Andosca stood up and talked about how cows are being used for energy harvesting and gave us the best quote: “You power the Moo-mometer with MEMS because cows get dirty.” Tech-Eye reporter Tamlin Magee loved that one too and plans to write a story on – perhaps cow-power is the next big thing!

MEMS Executive Congress Europe 2013The last panel of the day before the closing keynote was MEMS in medical with a focus on aging moderated by Frank Bartels, Founder (Bartels Mikrotechnik), President (IVAM). Panelists were:  Heribert Baldus, Principal Scientist – Personal Health Solutions, Philips Research; Jérémie Bouchaud, Senior Principal Analyst, MEMS and Sensors, IHS iSuppli; Kimmo Saarela, CEO, TreLab Oy; and Axel Sigmund, National Contact Point MTI/DW and Ambient Assisted Living Joint Programme, VDI/VDE Innovation + Technik GmbH. This was another diverse panel with varying views on how to address the medical and healthcare issues of the world’s aging population.

 When asked how MEMS is enabling a better quality of life with regard to prevention, monitoring, management, replacement and rehab I think Kimmo summed it up best when he said that with MEMS we can put so many things into a small form factor, which entices people to use our products. MEMS sensors allow us to collect raw data from so many sources. Data analysis is the key benefit and is their “value add” to the customer. But the key thing here is that power consumption and size really matter. Heribert added that MEMS is enabling an aging population to detect issues in their daily lives and manage their lives. I like to say it gives them their dignity back – and that is no trivial thing.

Jérémie spoke of some of the mass markets already present for MEMS in aging including sleep apnea disorders and oxygen therapy. There are also mass markets for MEMS medical applications that are in the hospital (not yet in the home) including disposable blood pressure monitors as well as dialysis and drug infusion applications. This kicked off a discussion about an aging population living at home which is becoming more of a critical issue in Europe, and a main focus of what Axel is addressing at VDI/VDE Innovation + Technik.

At the close, the panelists were asked what they saw as the future of medical – Heribert said he’d like to see more sensor integration, more intelligence and far less power. Jérémie said he sees a future for gas sensors analyzing the breath (and will not require FDA approval). Axel sees non-invasive diabetes monitoring as having the biggest impact; while Kimmo echoed Heribert and sees a future of more integrated solutions where biometric sensors will give more data and aid early detection and intervention. Frank agreed with Jérémie that gas sensors will be next once the pump issue is solved and that the time for microfluidics is near.

This final panel set things up perfectly for our closing keynote, Renzo dal Molin, Advanced Research Director, Cardiac Rhythm Management business unit, SORIN GROUP. Renzo gave the presentation “Vision for Implanted Medical Devices Healthcare Solutions and Technical Challenges,” which outlined the opportunity for implantable medical devices. He described in detail how

MEMS Executive Congress Europe 2013

the next generation of medical devices will come from miniaturization of devices, reduction of power consumption, and wireless capability and yes, even spoke of energy harvesting (you can guess whose ears perked at that statement). Renzo then highlighted how the BioMEMS market is expected to grow from $1.9 B in 2012 to $6.6 B in 2018 thanks to the inclusion of accelerometers in pacemakers and homecare monitors; MEMS sensors for glucose meter connected to smartphones; MEMS microphones for hearing aids as well as MEMS insulin pumps.

The audience was excited to discuss where Renzo saw the future of BioMEMS going, and where he felt the industry should focus moving forward. Renzo agreed that in the near future (once regulatory hurdles were overcome) patients will be able to monitor their implantable devices on their mobile devices. And he felt the next big thing will be biomarkers, as well as MEMS-enabled devices that could give an ECG will be revolutionary to the medical field.

MEMS Executive Congress Europe 2013And with that it was time to break and enjoy a fantastic evening at the Heineken Experience. We took some photographs throughout the day but by far my favorites are the ones we took at the brewery – you should definitely check them out. I would like to close this mega-long blog by thanking everyone who made this second-year MEMS Executive Congress Europe a great success from my fabulous MIG Team, to the MIG Governing Council, to the Congress EU Steering Committee, to the AMAZING sponsors (especially those top tier ones who are sponsoring all year long – we love you), the keynotes, the speakers, the attendees (especially the press who attended and those who have posted great stories – hooray!), our fantastic conference organizers at PMMI, and our sister conference folks at Smart Systems Integration. THANK YOU ALL.

Karen’s blog from MEMS Executive Congress: Part 1

There were many things that impressed me from hosting the second MEMS Executive Congress Europe – and it wasn’t the cold and snow (though it was chilly!). What struck me the most was how lively, engaged and intelligent the conversations were, not amongst the panelists but between the audience and the panelists. Often, Europeans can be conservative and reserved in conferences, but not this year In fact my favorite quote from one of the panelists was: “when I agreed to this join this panel I didn’t know I would be joining a religious war.”

MEMS Executive Congress Europe 2013The morning definitely didn’t start off with an aggressive tone as the elegant Ralf Schnupp, Vice President Segment Occupant Safety & Inertial Sensors, Continental served as our keynote. He focused his discussion on future trends in automotive with an overview of the megatrends affecting: safe mobility, clean power, intelligent driving, global mobility and most importantly, safety, with a goal of zero fatalities and accidents (WOW). He spoke of the challenges of complex sensor systems as well as the requirements of such systems. What stuck with me was his statement that “we don’t need more sensors, we need more robust, secure and safe MEMS/sensors.” For sensors I think he’s onto something (because it’s about the smart sensor integration and the software); although when I tried out that theory later that week at our sister-conference, Smart Systems Integration, I was completely shot down (ha!).

After Schnupp’s keynote came the consumer panel moderated very capably by Dave Thomas, Marketing Director, Etch Products, SPTS Technologies. Panelists included: Paul Buijs, General Manager, Bruco Integrated Circuits bv; Robin Heydon, Global Standards – Research and Innovation Group, CSR; and Joel Huloux, Director – Standardization and Industry Alliances, STMicroelectronics. You can probably tell from two of the four titles that the panel talked A LOT about standardization. And yes that was by design, as it’s an important topic that the MEMS industry has been working on and partnering with groups like MIPI Alliance (which Joel chairs).
MEMS Executive Congress Europe 2013

Joel brought a good perspective to the panel because he’s not a MEMS guy; he’s really an OEM/end-user that having spent over a decade with handset company Erikson (I want to say 20 years but don’t quote me) and is now with ST, because of the ST/Erickson joint venture. He said that MIPI aims to create specifications for mobile interfaces and recently became interested in MEMS (and joined an important partnership with MEMS Industry Group) because mobile devices add at least two new MEMS each year. True, but the question remains, what are you going to standardize? And with that question, thus opened a little bit of the holy war amongst the panel and the audience. Clearly it’s an important hot button issue.

When asked about the future of consumer electronics, the panelists all felt that its market strength would continue. Robin felt the most important impact on the world would be the Internet of things as well antenna switching (he does work for CSR after all). He also felt that the next move would be towards peripherals such as the smart watch – while Paul envisioned a future where we’d all have a “doctor in a watch” as the next killer app, enabled by MEMS.

Next up was the automotive panel moderated ably by Marc Osajda, Director, Pressure Sensor Business Unit, Freescale Semiconductor – Germany. With panelists: Frédéric Breussin, Business Unit Manager, MEMS & Sensors, Yole Développement; Pietro Perlo, Vice President Torino E-District, Interactive Fully Electrical Vehicles; and Jan Peter Stadler, Senior Vice President of Engineering Sensors, Automotive Electronics Division, Robert Bosch GmbH. What surprised me about this panel is how quickly the panelists started talking about electric bicycles (e-bikes). I actually had to check with Ralph Schnupp, who was sitting next to me, to confirm that was indeed what Pietro had started the panelists discussing.
MEMS Executive Congress Europe 2013

Marc quickly moved them back to automotive and it was actually quite comical to watch – Pietro and Jan Peter were sort of like the odd couple – both representing opposite sides of the spectrum of automotive. While Pietro focused on totally electric vehicles (including bikes!),Jan Peter averred that the automobile would evolve, but even by 2020 the majority of cars will still be run by combustible engines. Frédéric was well placed as a market analyst to give perspective on current uses of MEMS and sensors in applications such as night vision, heads up displays as well as efforts to reduce emissions, increase comfort and increase safety. What was also clear from all the panelists was that the consumer world is driving more and more of the automotive world; which is good for technology, but bad for pricing.

The best part of the panel was when Marc asked each panelist to describe what his car would look like in 2025. Frédéric said he’d finally give in and buy a hybrid, Jan-Peter said he wasn’t sure what kind of engine but he’d definitely want a car big enough to hold the wine he’d drive back from Romania and carry his e-bike to all the places he likes to use them (in the mountains). Lastly, Pietro stole the show when he said he’d be using a flying an electro-mobility flying car: “this is a possibility because we are MEMS!”

I’ll leave you hanging there, wanting to hear more of the excitement and challenging conversations at MEMS Executive Congress Europe 2013. A teaser: The next panel was MEMS in Energy, which discussed energy harvesting MEMS in depth, and as you can imagine, the opinions varied widely, to put it mildly. Soon, I’ll also describe to you the MEMS in Medical, focused on Aging panel which challenged us all to think more about quality of life issues and what more we can do with MEMS to enable a better world. So stay tuned, I’ll post my next blog soon.

Building a virtual gyro

Originally posted by Michael E Stanley of Freescale Semiconductor in The Embedded Beat on Mar 12, 2013

In Orientation Representations Part 1 and Part 2, we explore some of the mathematical ways to represent the orientation of an object. Now we’re going to apply that knowledge to build a virtual gyroscope using data from a 3-axis accelerometer and 3-axis magnetometer. Reasons you might want to do this include “cost” and “cost”. Cost #1 is financial. Gyros tend to be more expensive than the other two sensors. Eliminating them from the BOM is attractive for that reason.  Cost #2 is power. The power consumed by a typical accel/mag pair is significantly less than that consumed by a MEMS gyro. The downside of a virtual gyro is that it is sensitive to linear acceleration and uncorrected magnetic interference. If either of those is present, you probably still want a physical gyro.

So how do we go from orientation to angular rates? It’s conceptually easy if you step back and consider the problem from a high level. Angular rate can be defined as change in orientation per unit time. We already know lots of ways to model orientation. Figure out how to take the derivative of the orientation and we’re there!

In our prior postings, we’ve discussed a number of ways to represent orientation. For this discussion, we will use the basic rotation matrix. Jack B. Kuipers has a nice derivation of the derivative of direction cosine matrices in his “Quaternions and Rotation Sequences” text – one of my most used textbooks.  It makes a good starting point.  Paraphrasing his math:

Let:

  1. vf = some vector v measured in a fixed reference frame
  2. vb = same vector measured in a moving body frame
  3. RMt = rotation matrix which takes vf into vb
  4. ω = angular rate through the rotation

Then at any time t:

  1. vb= RMt vf

Differentiate both sides (use the chain rule on the RHS):

  1. dvb/dt  = (dRMt/dt) vf + RMt(dvf /dt)

Our restrictions on no linear acceleration or magnetic interference imply that:

  1. dvf/dt = 0

Then:

  1. dvb/dt  = (dRMt/dt) vf

We know that:

  1. vf = RMt-1 vb

Plugging this into (8) yields

  1. dvb/dt  = (dRMt/dt) RMt-1 vb

In a previous posting (Accelerometer placement – where and why) , we learned about the transport theorem, which describes the rate of change of a vector in a moving frame:

dvf/dt = dvb/dt – ω X vb

Those who take the time to check will note that we have inverted the polarity of the ω in Equation 11 from that shown in the prior posting.  In that case ω was the angular velocity of the body frame in the fixed reference frame.  Here we want it from the opposite perspective (which would match gyro outputs).

And again,

  1. dvf/dt = 0 so
  2. dvb/dt = ω X vb

Equating equations 10 and 13:

  1. ω X vb = (dRMt/dt) RMt-1vb
  2. ω X = (dRMt/dt) RMt-1

where:

  1. 0 z ωy
    ω X = ωz 0 x
    y ωx 0

Going back to the fundamentals in our first calculus course and using a one-sided approximation to the derivative:

  1. dRMt/dt = (1/Δt)(RMt+1 – RMt)

where Δt = the time between orientation samples

  1. ωb X = (1/Δt)(RMt+1 – RMt) RMt-1

Recall that for rotation matrices, the transpose is the same as the inverse:

  1. RMtT = RMt-1
  2. ωb X = (1/Δt)(RMt+1 – RMt) RMtT

Equation 15 is a truly elegant equation.  It shows that you can calculate angular rates based upon knowledge of only the last two orientations.  That makes perfect intuitive sense, and I’m ashamed when I think how long it took me to arrive at it the first time.

An alternate form that is even more attractive can be had by carrying out the multiplications on the RHS:

  1. ωb X = (1/Δt)(RMt+1 RMtT – RMt RMtT)
  2. ωb X = (1/Δt)(RMt+1 RMtT – I3×3)

For the sake of being explicit, let’s expand the terms.  A rotation matrix has dimensions 3×3.  So both left and right hand sides of Eqn. 22 have dimensions 3×3.

  1. (1/Δt)(RMt+1 RMtT – I3×3)  = (1/Δt) W
  1. 0 W1,2 W1,3
    W = RMt+1 RMtT – I3X3 = W2,1 0 W2,3
    W3,1 W3,2 0

The zero value diagonal elements in W result from small angle approximations since the diagonal terms on RMt+1 RMtT will be close to one, which will be canceled by the subtraction of the identity matrix.  Then:

  1. 0 z y 0 W1,2 W1,3
    ω X = z 0 x =  (1/Δt) W2,1 0 W2,3
    y x 0 W3,1 W3,2 0

and we have:

  1. ωx= (1/2Δt) (W3,2 – W2,3)
  2. ωy= (1/2Δt) (W1,3 - W3,1)
  3. ωz= (1/2Δt) (W2,1 - W1,2)

Once we have orientations, we’re in a position to compute corresponding angular rates with

  • One 3×3 matrix multiply operation
  • 3 scalar subtractions
  • 3 scalar multiplications

at time each point.  Sweet!

Some time ago, I ran a Matlab simulation to look at outputs of a gyro versus outputs from a “virtual gyro” based upon accelerometer/magnetometer readings.  After adjusting for gyro offset and scale factors, I got pretty good correlation, as can be seen in the figure below.

image001.gif

You will notice that we started with an assumption that we already know how to calculate orientation given accelerometer/magnetometer readings.  There are many ways to do this.  I can think of three off the top of my head:

  • Compute roll, pitch and yaw as described in Freescale AN4248.  Use those values to compute rotation matrices as described in Orientation Representations: Part 1.  This approach uses Euler angles, which I like to stay away from, but you could give it a go.
  • Use the Android getRotationMatrix [4] to compute rotation matrices directly.  This method uses a sequence of cross-products to arrive at the current orientation.
  • Use a solution to Wahba’s problem to compute the optimal rotation for each time point.  This is my personal favorite, but I think I’ll save further explanation for a future posting.

Whichever technique you use to compute orientations, you need to pay attention to a few details:

  • Remember that non-zero linear acceleration and/or uncorrected magnetic interference violate the physical assumptions behind the theory.
  • The expressions shown generally rely on a small angle assumption.  That is, the change in orientation from one time step to the next is relatively small.  You can encourage this by using a short sampling interval.  You should soon see an app note that my colleague Mark Pedley is working on that discards that assumption and deals with large angles directly.   I like the form I’ve shown here because it is more intuitive.
  • Noise in the accelerometer and magnetometer outputs will result in very visible noise in the virtual gyro output.  You will want to low pass filter your outputs prior to using them.  Mark will be providing an example implementation in his app note.

This is one of my favorite fusion problems.  There’s a certain beauty in the way that nature provides different perspectives of angular motion.  I hope you enjoy it also.

References

  1. Freescale Application Note Number AN4248: Implementing a Tilt-Compensated eCompass using Accelerometer and Magnetometer Sensors
  2. Orientation Representations: Part 1 blog posting on the Embedded Beat
  3. Orientation Representations: Part 2 blog posting on the Embedded Beat
  4. getRotationMatrix() function defined at http://developer.android.com/reference/android/hardware/SensorManager.htmlWikipedia entry for “Wahba’s problem”
  5. U.S. Patent Application 13/748381, SYSTEMS AND METHOD FOR GYROSCOPE CALIBRATION, Michael Stanley, Freescale Semiconductor