Why live timing is harder than it looks
Sub-second GPS, patchy trackside connectivity, and the tradeoffs behind Outlap's phone-based lap timer.
Lap timing sounds simple: cross a line, stamp a time, compute a delta. In practice, the constraints stack up fast once the “line” is a GPS coordinate and the “computer” is a phone in someone’s pocket at 140 km/h.
The core problem
Consumer GPS updates at 1–10 Hz and has meter-level jitter. At racing speeds, a single missed fix can shift a lap time by tens of milliseconds — enough to matter for a club-day leaderboard, even if it’s noise for anything safety-critical.
What we optimise for
- Consistent bias over absolute accuracy. A timer that’s 50ms slow every lap is more useful than one that’s accurate on average but noisy lap to lap.
- Graceful degradation. Trackside connectivity is unreliable — timing has to keep working offline and reconcile once a connection comes back.
- Cheap to deploy. No beacons, no transponders. A phone and a circuit map.
Where this is going
The next iteration folds in sector splits and a predicted-lap overlay, both covered in a future post.