Most people have never thought to count how often their leg bounces during a day. We hadn't either — until we built BinBot to measure it.
The scientific premise
Levine's research on Non-Exercise Activity Thermogenesis (NEAT) found that incidental motion like fidgeting can account for up to 350 kcal/day difference between people. Not formal exercise — the constant background micro-movement of someone who simply can't sit still. That number is the reason BinBot exists.
What happens under the hood
The core is CoreMotion sampled at 100Hz, passed through a 3-9Hz IIR bandpass filter. Walking cadence sits around 2Hz; most ambient vibration is above 10Hz. The 3-9Hz window isolates the frequency range where leg-bounce typically lives.
Above the filter, a five-stage rhythm classifier distinguishes three motion types: bounce, jump, and spin. Every session writes its count to HealthKit, so fidget data appears in the same Health app timeline as steps and heart rate. The app also supports Apple Watch, a home screen widget, and Live Activity for the lock screen.
Where it's imperfect
Walking cadence and car vibration occasionally land in the 3-9Hz range and trigger false positives. The filter reduces this but doesn't eliminate it. Watch-to-iPhone sync is currently one-way. Very low-amplitude fidgets may fall below the detection threshold.
We're naming these upfront because the gap between "interesting motion data" and "clinical accuracy" is real. BinBot is firmly in the first category.
Why free with ads
We wanted to see the real variance in fidget counts across different people more than we wanted subscription revenue. Free removes the barrier. The trade-off is ads.
Top comments (0)