In order to have a real impact, wearable fitness devices must make better use of data and integrate machine learning, according to Matt Landers (pictured), CEO of Stonecrysus, a start-up developing a device and app to address these current shortcomings.

Landers told Mobile World Live that data collection is generally not accurate enough on current wearable devices, and that the information that is collected is not then analysed sufficiently to be meaningful.

This is a problem with the recently-announced health apps from Apple and Google, Landers said, as they still rely on simple data, such as steps taken and food portion sizes, with little analysis.

However, Landers wasn’t completely dismissive of devices currently available: “The initial introduction of products is a great first step so people are aware of their activity,” he noted.

Stonecrysus uses machine learning and intelligent data collection to provide users with more accurate feedback and a better understanding of the effect their activity is having on their body.

While most activity trackers use accelerometers, the Stone has additional processors to improve accuracy of the data collection.

The Stone can be set up for different activities, meaning it can calculate the effect a movement on an individual depending on their activity. For example, running will have a different fitness result to cycling, yoga or weight lifting. The device also tracks sleep activity.

Users can input what they’ve eaten via the Android app, which is connected to the Stone via Bluetooth, using a simple graphical interface.

Intelligent learning algorithms then work out how food intake relates to the results of physical activity. Salt intake, for example, will lead to greater water retention, which will then partially offset weight loss achieved through exercise.

Users also enter information about their body type and fitness goals with progress tracked visually via the patented Fitness Arc, rather than using weight or calories burned, which the company has found to be less effective in terms of motivating users.

In total, the Stone device and app are able to take into account 35 different measurements which can be used to help users understand how to reach their desired fitness or health goals.

“Big data is kind of the next step,” Landers suggested.

Indeed, StoneCrysus enables users, as well as health professionals and fitness trainers, to access more detailed information and gain a deeper insight into the data via a dashboard product.

Stonecrysus recently completed a beta programme of its Stone wearable device and Android app and is now looking to secure $100,000 via a Kickstarter campaign to fund an initial production run of 5,000 units.

The company is working on an iOS version of the app and with one patent granted, has three pending, according to Landers.