·Insights

Race Intelligence

Every prediction below was produced during the race and validated against ground truth after. ML model validation — what the models saw vs. what actually happened.

Tire health over race

VER

LEC

NOR

Each driver falls toward 0 through the stint and rises after a pit — back to 1.0 on a fresh set, or lower if the tire fitted already has laps on it. Pit events shown as vertical markers.

Tire Degradation

Live

Tracks how much performance budget a tire has left on a 0–1 scale. Warns 3+ laps before the cliff that causes sudden lap-time loss — earlier than any broadcast graphic.

Algorithm

A Kalman-filter state-space model tracks each tyre's degradation trajectory lap by lap, calibrated per compound and circuit.

Key metrics

0.0 – 1.0Output scale

All circuitsCoverage

Per lapUpdate cadence

~0.5msInference

In your webhook payload

analytics.tireHealth = {
  tireHealth: 0.38,           // 0–1 performance index
  cliffLapPredicted: 27,      // predicted cliff lap
  cliffRisk: "THERMAL",       // NONE | GRAINING | THERMAL | BLISTERING
  degModeActive: "DEGRADING"  // WARM_UP | STABLE | DEGRADING | CLIFF
}

Your product could know all of this before the pit wall does.

Everything shown above — tire health, pit windows, safety car probability, championship points — delivered as structured JSON during the live race. Per driver, per lap, via webhook.

Live webhook delivery from $99/mo · Full analytics stack — talk to sales

Start building →Read the docs →Talk to us →