A coming wave of high-performance MEMS based inertial sensors will enable lots of new applications. We recently spoke with Dr. Andrei Shkel, a Professor at the University of California at Irvine, who has served as a Program Manager at DARPA's Microsystems Technology Office. Dr. Shkel is also the event chair and organizer for the upcoming International Symposium for Inertial Sensors and Systems (ISISS 2014) with participation from such companies as Northrop Grumman, Honeywell, Analog Devices, Qualtre, and Systron Donner. In this comprehensive interview, Dr. Shkel shares his views on the current R&D and commercial trends, provides an overview of relevant DARPA programs, and outlines his predictions for emerging applications enabled by the next generation of MEMS based inertial sensors.
MEMS Journal: For MEMS based inertial sensors, what are some of the main accomplishments in the R&D community in the past 2-3 years? Why were these milestones significant?
AMS: Over the last 2-3 years, we have seen a definite trend in aggressive miniaturization and development of wafer-level packaging technologies allowing to pack more inertial sensors (and non-inertial aiding sensors) in a small volume, from a single sensor to complete Inertial Measurement Units (IMUs) with six sensors (roll-pitch-yaw-x-y-z) to extended IMUs with 9 and more sensors (IMU + clock + magnetometer + thermal + altimeter). We used to have one or two companies with know-how in inertial sensors, now there are dozens of new companies entering the market.
These sensors are still very low-grade, they are "just good enough" for some basic consumer electronics applications, but developments are an indication of maturity of the MEMS community, availability of new foundries in US, Europe, and Asia, and wide-spread access to standard processes. There are also some interesting trends in miniaturization of high performance sensors.
MEMS Journal: Again, for MEMS based inertial sensors, what are people working on now in the research community? What big milestones do you anticipate this and next year?
AMS: The research community in MEMS inertial sensors is working on development of sensors which are more sensitive and very stable over a long period of operational time; the latter is a significant problem with currently available MEMS inertial sensors. This is almost a "religion" of the MEMS community -- we want these devices to preserve all the benefits of MEMS: small size, weight, and power plus cost (SWaP+C). No boutique manufacturing processes are appreciated; the main interest is in batch manufacturing techniques when thousands of devices are produced at the same time.
The community is exploring very different approaches: ingenious structural designs, MEMS with new high Q-factor materials, innovative precision fabrication, elegant wafer-level packaging, sophisticated on-chip calibration, and of course exploration of new physical principles.
We are now starting to see the first results in literature demonstrating gyroscopes with sub-degree/hour and accelerometers with sub mili-g in-run bias stability. In one to two years, these technologies will be a foundation for new companies producing a new generation of inertial sensors, and those it turn will enable new applications.
MEMS Journal: What are some of the most interesting research projects currently funded by DARPA and the other federal agencies? Why are these of interest?
AMS: Over the last 5 years, DARPA's Microsystems Technology Office has been developing a focused portfolio of programs, called "Microtechnology for Positioning, Navigation, and Timing (micro-PNT)". The portfolio is solving "an old problem with a modern technology" of knowing where you are and what time it is, in a very compact form factor, with high precision and all the time, regardless whether the GPS signal is available, jammed, or intentionally compromised.
The programs are very, very exciting with potentially a very consequential impact on warfighters. The micro-PNT program portfolio is listed on the DARPA website. Here, I won't be able to talk about all advances of the last five years, but let me highlight just a few revolutionary developments.
Microscale Rate Integrating Gyroscopes (MRIG) program is creating a new 3D micro-fabrication technique based on the glass-blowing paradigm. We may see very soon very high performance 3D gyroscopes that look like a micro wine glass made out of fused silica or other high-Q materials. Such structures may demonstrate a new operational modality, so-called Whole Angle Mode Operation, with a potential of precision navigation of fast spinning objects.
The second program I'd like to highlight is Primary and Secondary Calibration on Active Layer (PASCAL). PASCAL is developing self-calibration capabilities right on-chip with a goal to eliminate long-term bias and scale-factor drift of inertial sensors. This is a perfect example of a sophisticated micro-system within a micro-system, a sort of "lab-on-a-chip" where the "lab" provides the inertial calibration with large and heavy rate tables and massive shakers being replaced by microstructures.
The Timing and Inertial Measurement System (TIMU) program is trying to develop a technological foundation for a chip with some very real navigation capabilities which is of the size of an apple seed (this is a reduction by 4 orders magnitude of the conventional apple-sized IMUs).
A very exciting new program is Chip-Scale Combinatorial Atomic Navigator (C-SCAN) that combines inertial sensors with dissimilar, but complementary, physics of operation into a single microsystem. This program will co-integrate, in a very small volume, atomic and solid-state inertial sensors.
All these programs are solving a very real problem of navigation and guidance when GPS is not available. While at DARPA, I had a few short articles in the GPS World magazine discussing some activities of the micro-PNT program. The Office of Naval Research (ONR) also has a number of developments in inertial sensors technology.
MEMS Journal: What are your thoughts about using MEMS motion sensors for indoor positioning and navigation? How close are we to having commercial sensors for these applications that cost $1-2 or less?
AMS: The general problem of indoor positioning and navigation is a very difficult problem. It will be a while until we find a generic solution, and it’s unlikely that this solution will cost $1-2 or less. Currently, the long-term inertial navigation requires very expensive, heavy, and power hungry sensors. The problem will be likely segmented, i.e. the solutions will be application-specific. The general trend is for probabilistic navigation utilizing the inertial information along with signals of opportunity for recalibration. Some partial solutions are possible with Wi-Fi networks, distributed in-door radars, and other enabling technology. As the chip-scale inertial sensors and clocks are improved, there will be more interesting opportunities for indoor navigation, authentication, and fast relocking to GPS.
MEMS Journal: How do MEMS gyros currently compare with non-MEMS based technologies such as FOGs and RLGs?
AMS: Some high-end MEMS gyroscopes are approaching the performance of FOG and RLG, probably an order of magnitude lower sensitivity and two orders of magnitude lower long-term stability. However, the prices of such high-end MEMS are not significantly cheaper. Small SWaP advantages are considered to be a product differentiator, allowing the manufacturers to sell the units at the high cost. The main beneficiaries of high-performance MEMS sensors in military applications are small-diameter missiles, underwater navigators, portable North-finders, and UAV/UUV.
MEMS Journal: Is there a fundamental limit for MEMS inertial sensor performance due to the small proof mass in MEMS devices?
AMS: In general, the small and high-performance inertial instrument is a contradiction in terms. If you want a high inertial signal you would use a large inertial mass. The scaling in size is also not favorable for scaling of fabrication imperfections. On one side of the scale, we have SWaP+C story and an exciting opportunity for co-integration of MEMS with electronics; however, on the other side, we have all the disadvantages of scaling.
These trade-offs are very non-trivial, and future breakthroughs in microtechnology for positioning, navigation, and timing (PNT) will likely rely on yet-to-be-exploited physics, new materials, highly specialized fabrication technologies, and batch assembly techniques, selective wafer-level trimming and polishing, a combination of passive and active calibration techniques strategically implemented right on-chip. Introduction of innovative test technologies will also play an important role.
MEMS Journal: What are the highest performance MEMS inertial sensors (both accelerometer and gyro) that are currently available on the market?
AMS: The highest performance inertial sensors available on the market today are on the order of 1 degree perhour for gyros and milli-g for accelerometers (called, upper tactical grade). However, these devices are ITAR restricted and cannot be purchased for a general use. The consumer electronics grade inertial sensors are widely available. The higher-end consumer grade sensors are several orders of magnitude lower than what is possible with MEMS today and the higher-end pricing is ranging from about $100 per axis to approximately $1,500 for an IMU.
MEMS Journal: Which companies are the leaders in high-performance inertial sensors?
AMS: The definite leaders in high-performance MEMS inertial sensor components and inertial systems in the US are Honeywell, Northrop Grumman, and Systron Donner (BEI). Most of them don't sell components and only sell the complete Inertial Measurement Units (IMUs) and Inertial Navigation Systems (INSs). Most of their systems are ITAR restricted.
MEMS Journal: Which new commercial applications for MEMS based inertial sensors do you see emerging the next 2-3 years?
AMS: Some obvious applications we see today are in interactive consumer electronics (cell phones, games, image stabilization in small cameras, and recreational sports) and automotive safety systems (airbag deployment systems and advanced breaking systems). These applications are mainly using the threshold detection.
As the performance and stability improves, inertial sensors will be used increasingly more for quantitative analysis in more sophisticated applications. I see a growing interest in use of sensors for performance training, command and control centers for professional sports, augmentation of computer graphics for superimposing and display of information, and also in authentication. Additional applications for high-performance inertial sensors are oil and gas exploration, implantable prosthetics for restoration of damaged function of the vestibular system, and guided micro-surgery.
MEMS Journal: What are the most exciting and interesting motion sensor startup companies?
AMS: At the moment I don't see any truly captivating startup companies in this space. However, the landscape is changing very fast. Let's stand by.
This article is a part of MEMS Journal's ongoing market research project in the area of MEMS based inertial sensors and applications. If you would like to receive our comprehensive market research report on this topic, please contact Dr. Mike Pinelis at [email protected] for more information about rates and report contents.
Copyright 2014 MEMS Journal, Inc.