For Olympic swimmers, the blink of an eye can be the difference between first and forgotten. Everyone wants an edge, which is why elite athletes train relentlessly and why coaches push them hard. Alongside traditional techniques, wearable smart devices are now helping save those vital milliseconds.
You wouldn’t notice these devices: they’re smaller than a thumb and measure swimmers’ movements – information that can then be uploaded to the internet and used by coaches to improve performance, whether or not at an Olympic level.
They can be used in combination. The sensors have the ability to “talk” to each other, which means data from one device (such as a heart-rate monitor) can be combined with another device (such as a mobile phone).
In this way, important information (such as “I’m going into cardiac arrest”) can be sent to people who need to know (such as your physician).
In other words, the information goes to the right person, at the right time.
Smart device, smart data
These smart devices have two roles to play – collecting and understanding data, and transmitting this data – and NICTA is developing technology to help with both.
Transmission of data via this type of sensor is part of a new international standard called wireless body area networks (WBAN) currently being developed by the Institute for Electrical and Electronics Engineers.
It doesn’t come without challenges.
In radio design lingo, we say “the human body is a bad channel”. But why?
Well, imagine you’re driving down a freeway. It’s a flat, straight piece of road on which you can drive at high speed. That’s a good channel: the data (you and your car) are able to move at high speed.
Now, imagine for a moment your car leaves the freeway and you start driving on the dirt beside the road. The bumpy surface means you can’t drive so fast. That’s a poor channel: very little data are able to be transmitted, effectively meaning almost no communication.
When a person moves, the quality of the body’s channel changes. This is because the human body itself affects the radio signals used by the network to transmit data. Whenever a person wearing a smart device moves (this could be something as simple as breathing) the radio signal can be lost.
NICTA’s smart sensor devices can transmit at 12megabits per second on a good channel, which is ten times faster than Bluetooth. But on a bad channel a device might not be able to transmit at all.
Given the human body’s ability to interfere with the quality of communication channels, body area networks are not reliable. Of course, this could be fatal if the measurements are of critical physiological signs (such as the user having a heart attack).
To make the body area network reliable, we must be able to predict whether or not a channel will be good or bad at a particular point in time.
To do this, our research team asked office workers to wear hip, wrist and ankle sensors and used the data to monitor how the body affects communications.
We analysed the measurements using predictive models known as Markov models which allow us to predict whether or not a channel will be good or bad in the future.
The Markov Model allows us to transmit high-rate data – such as video content – when the channel is good. Conversely, we can keep low-rate data for later, when the channel is bad.
Swimming with sensors
The Australian Institute of Sport (AIS) tried the smart sensors with their team of elite swimmers, who wore the devices on their costumes and swam their normal set of laps.
The information gathered helped in several ways.
Each device has a microcomputer that processes the raw data into a series of laps and strokes. Activity-aware software on the device provides the coaches with feedback including information about the swimmers’ best lap times or the number of laps they have left before the end of their training.
In fact, our smart sensors count the number of laps, the time taken, stroke rate and type of stroke for each lap. This means an athlete can record their entire swim, stroke-by-stroke and view the data later on.
Old written training diaries can now be phased out in favour of automatically-generated, online data, accessible from anywhere in the world.
While an athlete is training for an event, a coach can look through the training database and ask: “How many strokes of butterfly did you swim since March? – a level of detail that simply would not have been possible with the old stopwatch-and-paper-based approach.
The difference is information. When a coach knows exactly how a swimmer performed, he or she can help to improve the swimmer’s performance.
Every little helps
The smart sensors are also part of a new health approach known as ehealth: the use of electronic data to improve health.
The data provided by such devices will help Olympians and everyday swimmers improve their performance and technique, making a difference when it really matters.