TyrePulse AI

Inspiration

In motorsport, data is abundant, but actionable intelligence is rare. We were inspired by the chaos of a practice session—where fuel burn, traffic, and track evolution all conspire to hide the true performance of a tyre. We asked: "What if we could separate true tyre degradation from everything else happening on track?"

What it does

TyrePulse AI ingests raw practice session telemetry (lap times, fuel loads, sector dynamics) and applies a deconvolution engine to remove "noise" (fuel effect, traffic, track evolution). It outputs a clean, isolated degradation curve per compound, revealing exactly when performance drops off the "cliff edge."

How we built it

We built a mathematical model to solve for latent vehicle state parameters.

The Model: We used a Bayesian state-space framework combined with ML regression to model the equation: LapTime_obs = Pace_base + fuel + traffic + track + Degradation_tyre + ε

The Stack: We designed a real-time analytics platform to visualize the "Clean Curve" alongside raw data, turning complex telemetry into a strategic dashboard.

Validation: We implemented a post-race validation loop to compare predictions against actual race results, continuously improving the model's accuracy.

Challenges we ran into

The biggest challenge was the data conflict. Standard timing telemetry conflates fuel mass loss (which makes the car faster) with tyre wear (which makes the car slower). Disentangling these signals mathematically required rigorous modeling of the physics of the car and the track, ensuring we didn't mistake a drying track for a dying tyre.

Accomplishments that we're proud of

We successfully quantified the "Signal vs. Noise" ratio in F1 data. We can now mathematically confirm that true degradation is hidden behind a wall of speed improvements, and our model consistently finds that wall. We turned a messy spreadsheet problem into a one-dashboard intelligence tool.

What we learned

We learned that race engineering is as much about data science as it is about intuition. The physics of motorsport (fuel burn, temperature, track grip) must be treated as deterministic variables, while tyre degradation is a stochastic process. Success requires merging these two worlds.

What's next for TyrePulse AI

We plan to integrate live weather telemetry and predictive pit window algorithms. We want to move from "How many laps are left?" to "What is the optimal lap to pit for the undercut?" based on real-time competitor data.