The University of Southern California has published the results of a new study that sought to find out whether wearable devices could anticipate conflict between couples before they occur. Using machine learning techniques, the researchers were able to establish a system that could apparently capture conflict episodes with an accuracy rate of up to 86 percent.
The majority of the research was conducted outside of the lab, according to a report from TechCrunch. Couples were given wearable sensors and smartphones to record data, and asked to complete a survey every hour recording their thoughts and feelings regarding their significant other.
The wearable devices kept track of the subjects’ body temperature, heart activity, and sweat levels. This data was then cross-referenced with audio recordings, which were analyzed to gain insight into the content and intensity of speech so that the team could determine whether or not conflict had emerged.
The researchers hope to continue to develop their machine learning algorithm, in the hopes that an improved version could spot the physiological signs of conflict up to five minutes ahead of time. This functionality could potentially be made available on commercially available wearables as a stand-alone piece of software.
At present, keeping fit remains the most common reason for people to don a wearable device, both in terms of dedicated fitness trackers like the Fitbit, and more general-purpose devices like the Apple Watch. However, various different research projects are demonstrating that this kind of technology could be used to provide all kinds of different health benefits.
In December 2016, a wearable that can counteract tremor caused by Parkinson’s disease was showcased on the BBC television program The Big Life Fix. Earlier this month, a campaign to fund a wearable that can apparently provide relief from menstrual pain launched on Indiegogo.
People weren’t as quick to jump on the smartwatch bandwagon as many manufacturers anticipated that they would be. However, wearables that provide tailored health services may prove more popular, if they can offer tangible improvements to their users’ quality of life.