Activity trackers on a smartphone or wearable device are increasingly being used by healthcare providers and insurance companies to reward active individuals for healthy behaviour or to monitor patients. But these devices can be easily deceived.
New research published in the journal PLOS ONE could help tackle this problem by identifying when the wearer is attempting to cheat.
Scientists at Northwestern Medicine and the Rehabilitation Institute of Chicago (RIC) have designed a way to train smartphone trackers to spot the difference between fake and real activity, Northwestern University reported last week. For example, the new algorithm is able to detect when the user shakes the device to make it seem like they have taken a walk, or when they pretend to be sitting while they are actually walking.
“As healthcare providers and insurance companies rely more on activity trackers, there is an imminent need to make these systems smarter against deceptive behaviour,” commented lead study author Sohrab Saeb, a postdoctoral fellow at the Center for Behavioral Intervention Technologies at Northwestern University Feinberg School of Medicine. “We’ve shown how to train systems to make sure data is authentic.”
The researchers showed that smartphones rigorously trained on normal and deceptive activity can identify deceptive behaviour and generalise it across individuals. In other words, if the tracker learns how one person cheats, it will recognise the same behaviour in someone else.
According to the university, while systems trained on normal activity data predicted true activity with 38% accuracy, training on the data gathered during deceptive behaviour increased their accuracy to 84%.
“Very few studies have tried to make activity tracking recognition robust against cheating,” said senior author Konrad Kording, a research scientist at RIC and an associate professor in physical medicine and rehabilitation at Feinberg. “This technology could have broad implications for companies that make activity trackers and insurance companies alike as they seek to more reliably record movement.”
The method is not completely foolproof, however. “If someone attaches an activity tracker to a dog, the system can’t recognise that,” Saeb said.