At present while smartwatches are able to recognise a limited number of particular activities, in order to do so they have to be programmed in advance.
The new algorithm has been designed to enable the smartwatch to discover activities as they happen and can even track sedentary activities.
Commenting Dr Hristijan Gjoreski of the University of Sussex said: "Current activity-recognition systems usually fail because they are limited to recognising a predefined set of activities.” According to Dr Gjoreski this new machine-learning approach, “detects new human activities as they happen in real time, and outperforms competing approaches."
"Traditional models ' cluster' together bursts of activity to estimate what a person has been doing, and for how long," he explained.
This approach will take continuous steps and cluster them into a walk, but doesn't account for pauses or interruptions in the activity.
The new algorithm has been designed to track ongoing activity and it is able to play much closer attention to transitioning i.e. stops and starts, as well as the activity itself. For example, it will take into account short pauses and will assume that the walk will continue, holding the data while it waits.
Dr Daniel Roggen, head of the Sensor Research Technology Group at the University of Sussex, added: "Future smartwatches will be able to better analyse and understand our activities by automatically discovering when we engage in some new type of activity. This new method for activity discovery paints a far richer, more accurate, picture of daily life.
"As well as for fitness and lifestyle trackers, this can be used in healthcare scenarios and in fields such as consumer behaviour research."
