Free your fitness data from yourself
Unfit Bits outlines everyday techniques for generating the fitness datasets of your choice, enabling you to qualify for insurance discounts without the lifestyle to match.
Why Unfit Bits?
It is increasingly assumed that fitness trackers provide an objective view of the activities of their wearer. The assumption is that a person’s acceleration data as interpreted by some fancy algorithms, gives a robust insight into the fitness, health and behavior of their body, and cuts through the blurry ambiguities of memory and perception. During the last year, data from a Fitbit tracker has been used as evidence in court both in a case about the impact of a workplace injury on a worker’s health and more recently as evidence of a rape. How these early examples play out, will reveal how tight the relationship between activity data and behavior of the wearer is assumed to be.
In the near future, the choice not to log and share your fitness data is also likely to come at an expense. Privacy is set to be a costly luxury and already many insurance companies and banks are offering financial incentives to clients who agree to wear a freely provided tracker, but only if they log a certain level of activity. John Hancock, Alpha Bank and Oscar have all launched these sorts of programs this year. Large corporations are also encouraging their employees to wear trackers, last year BP distributed 25000 of these devices to their North American personnel and this year Target is following suit by offering them to their 335,000 employees in order to lower insurance premiums. In these cases, it is the employees that bear all the risk in terms of their privacy as they give over a very personal dataset to what remains an obscure data marketplace.
Yet anyone who has actually owned and worn a tracking device would have noticed that, their tracker’s algorithmic interpretation of their activity remains pretty crude. Leave your tracker in the washing machine and you’ll do your daily quota of steps half way through the spin cycle. Bike across town, and your tracker won’t see this as activity unless you manually log it. The perceived gravitas of this data in the context of court and insurance premiums overlooks its infidelity to the wearer’s activities.
This is why we’ve made Unfit Bits, an initiative that invites you to free your fitness from yourself. The project investigates a series of fitness tracker spoofs – ways to fool your tracker into logging steps you haven’t actually taken so that you too can qualify for insurance discounts and financial incentives without the lifestyle to match.
We attach trackers to metronomes, dogs, car wheels and branches moving in the wind and watch our faux step count tick over.
Through this research we also found that common trackers only give their users access to the algorithmic interpretation of their data in the form of steps and distances and not to the actual sensor data itself. This means that for the first time, the corporations behind this hardware have access to higher resolution data about the body of the wearer than the wearer themselves. The opacity of both these devices and their algorithms, also limits the capacity of the user to contest or corroborate the functioning of their device.
Using an open source accelerometer, we also compared the sensor data generated from our tracker solutions with the steps logged by a Fitbit and Jawbone. This algorithm audit showed that the Fitbit systems typically reports higher step counts than the Jawbone.
Introducing new break through exercise technologies
At Unfit Bits, we are excited to introduce our new range of desktop step machines. We have used our research to develop these simple desktop devices that enable you to step your way to freedom no matter what your lifestyle is like.
Made in collaboration with Surya Mattu.