Kiwi



The Perfect Wearable

In a short period of time, the consumer motion and activity recognition industry has bloomed, with devices flowering in almost every industry with numerous applications. As a result, the market is highly fragmented, with fitness trackers, smartwatches, gaming technology, VR headsets, and specialty gesture recognition devices for sports.

It can get overwhelming. Will you wear one device for sports, one for your health, one for payment, and one for communication? Or will there be a “super-wearable” that does it all?

 

And that’s what this post is about - What would a perfect wearable look like? Is it even feasible in the short-to-medium term? I’ve put together a wish-list of activity tracking features and will then examine the feasibility of those features.

 

One device, every motion

A device that tracks not just your steps, calories, sleep, heart rate, weight-lifting reps, sports movements, but more mundane and common activities like eating, driving, walking, sitting or watching TV.


For the next iteration of trackers, you may just download software patches for different activities, depending on your personal needs. An athlete’s requirements would naturally differ from that of the white collar worker.


Challenge

Lack of research - Because motion technology is still in its early stages, the industry lacks technical expertise and research. There are little to no common standards, best practices, or templates.


You can see the difference in the number of research papers published on IEEE, with only 86,100 for gesture detection and motion tracking as compared to 111,300,000 for voice recognition.  


As a result, every new motion technology device or application has to be built from the ground up, resulting in a long development timeline before it gets to market.


Pinpoint accuracy

Flawless and seamless motion recognition software is non-negotiable if wearables are to become mainstream. Whether it’s your caloric expenditure or a hand movement that switches radio stations in your car, each must be detected consistently and persistently.


While commercial activity trackers are able to provide data that’s good enough to see general trends, they show significant variations in accuracy, especially when compared to speciality devices.


Challenge:

Individual variation - Every person is different, but every wearable is the same. As a result, it’s difficult to build a wearable that caters to the infinite variations of the human race. For example, wrists that are too big, too small, too hairy or too tattooed may result in errors in the data.


However, as companies collect more data, this problem should diminish, as algorithms take a personalised approach. This may result in different versions of wearables based on your body type. It may look similar to the smartphone model, with devices of different shapes and sizes, as per consumer demand.  


Size won’t matter

The device will be small and comfortable enough to be worn the entire day. There may be a shift from the wristband activity trackers to ones that can be worn anywhere on the body, maybe even a tattoo, or an adjustable elastic band.


Challenge:

Sensors need to get cheaper. When sensors get cheaper, manufacturers are able to devote more resources on design and form factors. Although accelerometers and gyroscopes have seen a dramatic fall in price over the past few years, prices still need to fall further for most sensors to enable integration into one device.


Life-changing insights

How valuable would it be to get rapid, relevant and helpful data daily on methods to improve your health?

 

Studies report that many users of fitness trackers stop using them after a few months. The reason is simple. The trackers often don’t provide users a compelling reason to keep using them.

 

It’s not enough to know how many calories you’ve burnt, or hours you’ve slept. To elevate themselves from want-have to must-have, devices must deliver information that makes a tangible difference to your life.

 

Maybe you’re sleeping in an awkward position at night, have bad posture, or aren’t lifting properly at the gym. Your wearable would be able to tell you that you should sleep on your back, to extend your arms two inches further while lifting weights, or lean your head back and straighten your back.

 

Challenge:

Sensor integration – Sensor fusion has to improve to allow multiple sensors to communicate and talk to each other. The number of sensors being integrated into devices is rapidly increasing, and the streams of data need to be combined to provide intelligent feedback. Say, sensors in your bed interfacing with sensors that track your body movements to determine an awkward sleeping position and diagnose it before further harm can be done.