Reconstruct trips to the second
Trip traces built from vehicle-native samples: the route on a map, colored by what happened on it — harsh brakes, harsh acceleration, speeding segments — with crash-proxy incidents timestamped and located.
The problems we remove
He-said-she-said incidents
After an event, you get conflicting accounts and no physics. Who braked, when, from what speed — unknowable without data.
Objective reconstruction
Crash-proxy incidents record decel magnitude, speed before and after, severity band, and GPS. The trip trace shows the road context around the event.
Aggressive driving you can't see
Speeding and harsh maneuvers happen between stops. Managers only learn about patterns when something breaks — or someone gets hurt.
Event-mapped trips
Every harsh brake and accel is a marker on the trip map with its exact rate and speed. Recurring red segments on the same corridor reveal habits, not one-offs.
Sparse data, wrong conclusions
Phone-based detection fires on potholes and dropped phones. Infrequent polling misses the event entirely.
Vehicle-native samples
Speed and position come from the vehicle's own telemetry at up to 30-second cadence — no sensor fusion guesswork, and detection thresholds tuned for both dense and sparse windows.
What you can build on day one
GPS trip traces
Any trip renders on OpenStreetMap with segments colored: normal (blue), harsh brake (red), harsh accel (amber), speeding ≥115 km/h (dark red).
Event popups
Click a brake marker to see the decel rate in km/h/s, speed at that moment, and timestamp — the evidence trail.
Incident log
90-day incident table per vehicle: time, type, severity, decel, and speed transition. Webhooks push each one live.
Trip segmentation
Trips auto-segmented from drive sessions with duration, distance, avg/max speed, and sample counts — confidence-tagged for provenance.
Know what happened on the road
Trip reconstruction is live in the fleet console today — open any vehicle, expand a trip, see the map.