Greg Blackman investigates the latest miniature and wearable sensing technology, including 'smart' textiles incorporating fibre optic sensors to give data on movement
Miniature optical sensors are set to permeate most aspects of our lives, from the cars we drive to monitoring our health. If you want to know the chemical composition of the food you’re about to eat, the smartphones of the future will contain sensing technology to tell you.
Hamamatsu won a Prism Award at Photonics West in San Francisco at the beginning of February for a miniature spectrometer that can be connected to mobile devices like smartphones and tablet computers. It uses micro-opto-electro-mechanical system (MOEMS) technology to shrink the spectrometer to an instrument the size of a fingertip and weighing 5g.
Craige Palmer, general sales manager at Hamamatsu Photonics UK, noted that one of the largest markets for optical sensing technology is healthcare, adding that the miniaturisation of sensors and the development of portable instruments is moving medical testing from the hospital to the home. He said that this is being enabled by the ability to integrate filters, detectors and electronics into very small packaging.
In terms of the consumer market, sport is one area that has embraced smart sensing technology. There are devices available that record how heart rate and blood-oxygen content change during exercise – gadgets from companies like Masimo which released a fingertip pulse oximeter at the Consumer Electronics Show in Las Vegas at the beginning of the year. Masimo acquired the services of Stig Severinsen, four-time world champion freediver, and Dotsie Bausch, seven-time US cycling national champion, to demonstrate its MightySat pulse oximeter during CES. Severinsen held his breath inside a water tank during the show, (he has the world record of 22 minutes for a breath-hold), while Bausch went through her Olympic-level training, both connected to the pulse oximeter so that visitors could see how oxygen saturation of the two athletes’ blood altered.
The MightySat device attaches to the finger to give oxygen saturation, pulse rate and perfusion index, a relative assessment of pulse strength, and is based on the Masimo’s pulse oximeter technology used in hospitals. The instrument incorporates an optical detector to monitor the absorption of red and infrared light by the blood. Algorithms correct for any noise in the readings created by movement, necessary to make the technology robust enough for use in sport.
Also launched at CES was XelfleX from Cambridge Consultants, a ‘smart’ textile containing fibre optic sensors that gives data on the wearer’s movement. One of the main applications for this technology is for fitness and sports coaching, to help perfect a tennis serve or golf swing, for example. Other potential uses include those in physiotherapy to help patients recover from injury or even in motion capture in film making.
‘XelfleX came from the intersection of the areas we work in, namely high-performance fibre optic sensors for the oil and gas industry and consumer wearable sensors,’ explained Cambridge Consultants’ Martin Brock, the inventor of XelfleX.
‘We’ve noticed that a lot of existing wearable technologies are quite bulky and inelegant,’ Brock said. ‘What would be really nice would be clothing that was inherently smart, rather than needing to have an additional sensor attached. We’re trying to go beyond just fitness tracking or measuring how many calories you’ve burnt to really giving feedback on sports techniques. We’re looking to measure body motion and posture to help with technique in applications like tennis, golf or skiing, where billions of dollars a year are spent on equipment and coaching.’
XelfleX operates via optical time-domain reflectometry. It works by sensing the scattering properties of light from a laser diode travelling down a plastic optical fibre – the fact that the fibre is plastic makes the device cheap, safe, robust and washable.
The fibre is 100-200µm in diameter and is flexible enough to be integrated into a textile. The fibre crosses multiple regions where the garment flexes due to joint motion, for example at the elbow, wrist and shoulders. The garment has to be reasonably close-fitting, explained Brock, in order to stretch and flex in response to the wearer’s movement.
At each joint region, the fibre is bent in a number of loops. At the bend region some of the light gets reflected back towards the source and some scattered out, so that beyond the bend there is less light travelling down the fibre – the tighter the bend, the greater the amount of backscatter.
‘We can make multiple measurements by sending a short pulse of light down an optical fibre with multiple sensors attached in series along it,’ explained Brock. ‘Each time a pulse of light reaches one of the bending regions, some of that light gets reflected back to the start. Optical time-domain reflectometry works like a radar system – one pulse turns into a series of pulses returning separated in time by the extra delay that the light takes to propagate from one sensor to the next. From each reflection, we measure how strongly the fibre is bent and from that we can infer how much the joint has bent at that location.’
Each sensor gives one degree of freedom, one angle. To measure two angles – whether your wrist is flexing up and down or left and right, for example – would require two sensors. One fibre can accommodate around five to ten sensors (a shoulder might require between three and five sensors), and multiple fibres could radiate from a single electronics module, Brock stated.
‘The innovation in the patent we’ve applied for is how to do this in a low-cost way,’ Brock commented. ‘You can get lab instruments that can make these measurements, but they typically cost tens of thousands of dollars. Because we’ve introduced these regions of fibre with extra bends in them, we can optimise our measurement technique to operate at a very high speed but only on selected bits of the fibre. So, rather than needing to measure the whole fibre and its scattering properties, we’re homing in on just those five or ten regions we’re interested in. That reduces the cost and means we can make measurements quickly.’
A typical lab instrument might make a measurement a few times a second; XelfleX can record data points 100 times per second, and, according to Brock, could go faster.
Brock put the cost of a device with 10 sensors on one fibre at around $50. This one fibre could be woven into a shirt to generate basic sensing data on all the joints on both of the arms. ‘If you wanted more information you would need two fibres connected back to one electronics module,’ he said. ‘That would up the cost, but give higher quality information. Those are trade-offs we’re exploring with a number of companies that are interested in exploiting this technology.’
Very precise measurements are not needed to give useful feedback on sporting technique; it’s often more in the timing. For improving a tennis serve, for example, what matters is the speed of the elbow and wrist and how synchronised they are, more than the exact number of degrees the wrist is bent through the action, Brock explained. ‘You can get quite a lot of useful feedback if you develop the right algorithms without needing very high precision on the measurements, and that’s where we’re aiming with the basic system. We only need to measure strain to a few tenth of a per cent to get a good measure of how much a garment has bent.’
Brock added: ‘We can see both a technical way of making a system – and a commercial opportunity – for something that is a bit more expensive that gives a similar level of accuracy to a body motion capture system used for film making or gait analysis in a lab environment using cameras. That’s a lower volume, higher value application. There are medical physiotherapy applications with a similar price point and accuracy requirement.’
Cambridge Consultants is in licensing discussions with major sporting goods companies. ‘The differentiating factor of XelfleX is that, because it is a plastic thread, it really can be embedded into garments in a way that’s quite different to what’s been possible up until now with electronic sensors,’ Brock concluded. ‘It opens up this whole idea of smart clothing, where, other than a credit card-sized electronics module in a pocket, it is a very wearable garment; it’s not compromised by the technology. A lot of wearables at the moment are designed and aimed at techies rather than as wearable clothing.’