Despite lower than expected sales, wearables continue to be a growing market segment. Consumers spent over $2 billion on wearables in 2015 and are increasingly demanding better performance. One emerging application area for wearables, accurate biometric sensing, is highly complex and requires sophisticated solutions to provide accurate measurements during regular use, as well as during high-intensity workouts. The best performing solutions on the market are designed using a systems approach where both hardware and software are tightly integrated. Valencell is one of the developers of biometric sensing module technology for wearable devices. The company has raised a total of $24.5 million in venture funding to date, including $11 million in March of last year. We spoke with Ryan Kraudel, VP of Marketing at Valencell, to get the latest trends for biometric sensing in wearables including emerging applications and competitive dynamics.
MEMS Journal: Can you describe Valencell’s sensor solution?
Ryan Kraudel: From a hardware persepective, our sensor system includes LED emitter/detector sensor electronics, optical lenses, patented optomechanical designs, accelerometer to track activity and provide motion noise reference, and a low power microcontroller. On the software side it includes firmware and algorithms to remove noise during heavy activity and produce heart rate and other biometric measures (HRV, cardiac efficiency, blood pressure, etc). Our sensor solution comes in two varieties -- one for earbuds and one for wrist, arm and other body locations. Both varieties include: (1) an optical emitter/detector sensor electronics, (2) active signal characterization for signal processing, (3) low-power accelerometer to track activity, and (4) light-guiding optomechanics to provide optical coupling to the user’s skin.
MEMS Journal: Do you use standard components or custom hardware for your systems?
Ryan Kraudel: We use standard off-the-shelf components for our all our sensor systems.
MEMS Journal: What projects are you currently focused on and what are your plans for 2017?
Ryan Kraudel: We are continuing to develop new capabilities that will support more advanced user experiences -- for example, sleep metrics, training load, workout capacity, biofeedback training, and cardiac event identification, and monitoring. For 2017, we are very excited about measuring blood pressure with our standard sensor hardware that can be integrated into wearables and personal health devices. Some other key focus areas for us in 2017 include making our sensor modules smaller, more robust, and more power-efficient, while maintaining high levels of accuracy.
MEMS Journal: What are the main market opportunities for your technology and solutions?
Ryan Kraudel: Most of our business is in consumer wearables, but we are also seeing significant traction in personal health and medical solutions. We are getting an increasing number of inquiries about gaming, augmented reality (AR), virtual reality (VR), industrial safety, as well as military and first responder solutions.
MEMS Journal: Who are some of your customers and what do you provide to them?
Ryan Kraudel: Our biometric sensor technology is integrated into wearables and hearables from Bose, Sony, LG, Jabra, Samsung, Intel, and many more. We provide hardware and software soltions for these companies, as well as product design guidance, product testing support from our biometrics lab, automated manufacturing tests and quality control, and marketing support.
MEMS Journal: How do you typically work with customers?
Ryan Kraudel: We provide access to its technology in three different ways: (1) sensor modules that are complete biometric sensor systems with hardware, optomechanical design, firmware, and algorithms (no licensing), (2) product licensing that is typically utilized for devices that require non-standard integration of our technology (e.g. due to specific form factor requirements), and (3) patent licensing, which provides access to our inventions and patent portfolio excluding our technology products, services, or support.
MEMS Journal: What is your preferred engagement model and why?
Ryan Kraudel: Our preferred engagement model is to provide our sensor system, because it incorporates all of our proven technology in a pre-packaged module. This helps our customers get to the market rapidly and to scale up production faster than competing products.
MEMS Journal: How does your biometric sensing technology work? Can you describe how measurements are taken, processed, and analyzed?
Ryan Kraudel: Our technology uses a methodology called photophlethysmography (PPG) to measure heart rate and other biometrics. PPG is a technical term for shining light into the skin and measuring the amount of light that is scattered by blood flow. That is an oversimplification, but PPG is based on the fact that light entering the body will scatter in a predictable manner as the blood flow dynamics change, resulting from changes in blood pulse rates (heart rate) or in blood volume (cardiac output).
PPG-based devices shine visible or infared light into the body and use a photodetector to capture the light refracted from the user of the device and translates those signals into ones and zeros that can be calculated into meaningful heart rate data. An accelerometer measures motion and is used in combination with the photodetector signal as inputs into motion-tolerant PPG algorithms. The algorithms process the signals from the digital signal processor (DSP) and the accelerometer into motion-tolerant heart rate data, but can also calculate additional biometrics such as VO2, calories burned, R-R interval, heart rate variability, blood metabolite concentrations, blood oxygen levels, and even blood pressure. We have an active blog where we discuss biometric sensing and address common questions on heart rate monitoring.
MEMS Journal: What are the main limitations of PPG sensing?
Ryan Kraudel: PPG sensors have come a long way in terms of accuracy in the last five years, but still have their challenges in a few areas, such as motion-tolerance (PPG sensors are sensitive to optical noise, i.e. signals coming into the photodetector that are not relections of blood flow, particularly motion noise), skin tone (light is absorbed differently by different skin tones, which impacts accuracy in PPG sensors across diverse populations), the “cross-over” problem (PPG sensors can in some cases mistake step rate for heart rate), sensor location (the location of the PPG sensor on the body makes a huge difference), and blood perfusion (some people have low blood perfusion, meaning their blood has trouble certain extremities with enough consistent volume for accurate measurement).
While these issues remain for many PPG sensors, some of the most advanced sensor systems on the market today have solved these issues. People can get more detail on these challenges in our blog post on this topic.
MEMS Journal: What are the alternatives to PPG and how do they compare?
Ryan Kraudel: There are two primary alternatives for measuring biometrics in wearable devices, both of which are based on electrical methodologies: 1) bioimpedence and 2) electrocardiogram (ECG). Bioimpedence measures the resistance of body tissue to electrical signals created by your heart. However, bioimpedence has significant problems with motion noise as well, which makes measuring biometrics during activity or exercise very challenging. ECG measures the electrical signals produced by the heart. The most common form of ECG in wearables is a heart rate chest strap that come with many sports watches. However, many people find them uncomfortable and almost all the major chest strap vendors (Garmin, Suunto, Polar, etc.) are now using PPG in their latest wearables.
MEMS Journal: What are the main differentiators for your technology?
Ryan Kraudel: One differentiator is something called active signal characterization, which enables wearable devices to be highly motion-tolerant and measure biometrics accurately even when the user wearing the device is doing vigorous activity.
MEMS Journal: Which parameters can you measure with your sensors?
Ryan Kraudel: We can accurately measure continous heart rate, VO2, caloric burn rate, and total calories, R-R interval (RRI), breathing/respiration rate, activity recognition (running/lifestyle), step rate & total steps, pace and distance (walking or running), cycling cadence, signal quality, and off wrist/out of ear detection. Using these measuremnts, we can derive the following metrics: stress, training effect, HR zone, cardiac efficiency, HR recovery, resting HR, VO2 max, sleep metrics, and heart rate variability (HRV).
MEMS Journal: How can you determine the accuracy of the sensor measurements?
Ryan Kraudel: One way we ensure measurment accuracy is to provide our customers with a real-time signal quality measurement alongside the heart rate and other biometrics that we feed into the host processor. This gives our customers a “confidence score” for the quality of the data, so they can set thresholds under which they can choose to eliminate or separate data for a better user experience. For example, if the user of the device is not wearing the device properly or the device slipped out of position during exercise, the device or app where the data is shown can signal to the user that it is not tracking their biometrics appropriately and ask them to adjust the fit of the device.
MEMS Journal: How do you test and verify new products? Can you describe some of your test protocols?
Ryan Kraudel: We operate a biometrics testing facility at our headquarters in Raleigh, North Carolina. Led by exercise physiologist Dr. Chris Eschbach, the lab puts our customers’ product prototypes through a battery of testing protocols and activity sets. We then analyze and validate the data to determine real-world accuracy and identify improvements to product design and performance. We don’t stop until we demonstrate the product will deliver highly accurate biometric measurements on anyone, doing anything, at any time.
MEMS Journal: Who are your main competitors? How do your solutions match up against those of your competitors?
Ryan Kraudel: Our competitors are primarily silicon vendors and algorithm development shops. We strongly believes that biometric sensor systems must be designed and built to work as a system -- hardware, software, and testing are all critical components to making a highly accurate biometric wearable. Trying to piece together solutions in each of those three areas generally doesn’t lead to an accurate solution. There are other competitors in this market, including Philips, LifeQ, and Lifebeam. However, our technology is used in more biometric wearables than any other company in the world.
MEMS Journal: Are you working on any emerging applications for biometric sensing, such as virtual reality?
Ryan Kraudel: We’re seeing early interest for VR and AR applications of biometric sensor technology. It’s a natural fit both from a form factor perspective, because most of these devices are worn on the head which is a good location to measure biometrics, and also from a use-case perspective to understand how a person’s body is reacting to the experience.
MEMS Journal: Do you use the same fundamental sensor technology for medical and consumer applications? Are there any differences?
Ryan Kraudel: Yes, the same fundamental sensor technology is used for consumer and medical applications. We are in the early stages of working healthcare/medical device companies, so we can’t talk publicly about our work in detail at this point.
MEMS Journal: What are some of the military and first responder applications that you have worked on?
Ryan Kraudel: We’ve worked with one of the largest defense contractors in the world to integrate our biometric technology into radio earbuds worn by soldiers and first responders in the field. Again, it’s a natural fit for our technology because users already wear the radio earbuds and have a need for real-time, accurate biometric monitoring.
MEMS Journal: Are you doing anything with sensor fusion? If yes, what type of sensor data are you analyzing and what type of insights do you provide?
Ryan Kraudel: Because our standard sensor systems include optical sensors and accelerometers, we are by default using at least some sensor fusion. Our R&D teams are continually exploring how other types of sensors can contribute to more compelling user experiences with biometric wearables. For example, we recently completed a collaboration with ST that integrated our sensor system into the ST Micro SensorTile, which is a development kit that includes a gyroscope, magnetometer, digital microphone, and temperature sensor. The combined platform provides a comprehensive wearables development kit with every sensor you could want in a wearable device.
MEMS Journal: You have developed an ecosystem of chip suppliers, contract manufactures, and others. Can you talk about how you are using this ecosystem?
Ryan Kraudel: Manufacturing biometric wearables at scale is a different animal than building any other wearable device, because optics are involved. We set up a certification program for our supply and manufacturing partners to ensure the best outcome for our customers. The Valencell Certification Program ensures quality, repeatability and worker safety in the product manufacturing and quality control process. Our program certifies the engineering and manufacturing capabilities, equipment, and working environments of contract manufacturers (CMs) and original design manufacturers (ODMs) that build Valencell-powered products. This includes regular site visits to validate that certification requirements are met now and in the future.
This article is a part of MEMS Journal's ongoing market research project in the area of wearable and biometric sensors. If you would like to receive our comprehensive market research report on this topic, please contact Dr. Mike Pinelis at firstname.lastname@example.org for more information about rates and report contents.
Copyright 2017 MEMS Journal, Inc.