PitWall

F1 24 telemetry capture, storage, and analysis tool built in Rust. Designed for indie sim racers and amateurs who want lossless, high-frequency telemetry data they can query, replay, and analyze — with a live web-based dashboard.

What It Does

Connects to F1 24's UDP telemetry stream (port 20777)
Detects session start/end automatically
Records every telemetry sample (all Tier 1 + Tier 2 channels) at game tick rate
Stores data losslessly in compressed Parquet files (ZSTD)
Reconstructs and replays any lap frame-by-frame
Exports individual laps to standalone Parquet for DuckDB/pandas analytics
Optionally records raw UDP packets for deterministic replay during development
Live web UI — real-time F1-style dashboard via WebSocket

Live UI

A React-based dashboard that connects to the backend via WebSocket and displays real-time telemetry in an F1 broadcast style.

Features:

Speed / RPM / Gear display with animated RPM bar
Throttle and brake input bars
ERS energy level, deploy mode, DRS status, fuel remaining
Tyre temperature grid with color-coded heat indicators (blue → green → yellow → red)
Live lap timing table with sectors that fill in as they complete
Purple highlighting on fastest lap time and individual fastest sectors (live, not just after session)
Session header with track name, session type, live/offline indicator, session clock

Running the UI:

# Terminal 1: Start the backend (includes WebSocket server)
cargo run -- live f1

# Terminal 2: Start the React dev server
cd pitwall-ui
npm run dev

Open http://localhost:5173 — it auto-connects to ws://localhost:8765/ws.

For testing without the game:

# Terminal 1: Backend in dry-run mode
cargo run -- live f1 --dry-run

# Terminal 2: Simulate recorded packets
cargo run -- sim pitwall_data/recordings/<session_id>.pitraw

# Terminal 3: React UI
cd pitwall-ui && npm run dev

Option	Default	Description
`--ws-port`	8765	WebSocket server port for the live UI

Telemetry Channels Captured

Driver Inputs: throttle, brake, steer, clutch, gear

Vehicle State: speed, RPM, DRS status, DRS allowed, ERS energy, ERS deploy mode, fuel remaining

Position & Motion: world XYZ, velocity XYZ, yaw, pitch, roll

G-Forces: lateral, longitudinal, vertical

Tyre Temps: surface + inner for all 4 wheels (FL, FR, RL, RR)

Brake Temps: all 4 wheels

Suspension: position, velocity, acceleration for all 4 wheels

Wheel Speeds: all 4 wheels

Storage Layout

pitwall_data/
├── sessions/
│   └── {year}/
│       └── {track}/
│           └── {session_id}/
│               ├── metadata.json
│               ├── telemetry.parquet
│               ├── laps.parquet
│               └── events.parquet
└── recordings/
    └── {session_id}.pitraw

Parquet files use ZSTD compression
A 60-minute race produces ~10-50 MB of telemetry
Metadata is JSON for easy inspection
Raw recordings (.pitraw) are only created with --raw flag

Installation

# From the pitwall directory
cargo install --path .

# Or run directly without installing
cargo run -- <command>

Requires Rust 1.70+.

For the live UI:

cd pitwall-ui
npm install

Requires Node.js 18+.

Commands

`pitwall live f1`

Capture live telemetry from F1 24 with the WebSocket server for the live UI.

pitwall live f1 [OPTIONS]

Option	Default	Description
`-p, --port`	20777	UDP port to listen on
`-o, --output`	./pitwall_data	Storage directory
`--raw`	off	Also record raw UDP packets (.pitraw) for simulation
`--dry-run`	off	Process packets without writing to disk
`--ws-port`	8765	WebSocket port for the live UI

Examples:

# Standard capture with live UI
pitwall live f1

# Capture with raw recording for later simulation
pitwall live f1 --raw

# Listen on a custom port
pitwall live f1 --port 20778

# Process packets without storing (for testing with sim)
pitwall live f1 --dry-run

# Use a different WebSocket port
pitwall live f1 --ws-port 9000

Terminal output during capture:

🏎️  PitWall Live Mode → F1 24 UDP
   Port: 20777
   Live UI: ws://localhost:8765/ws
   Storage: ./pitwall_data
   Press Ctrl+C to stop

[INFO] Listening on 0.0.0.0:20777
[INFO] Live UI WebSocket server on ws://localhost:8765/ws
[INFO] First packet received (1349 bytes) - connection established
[INFO] Recording: track=11 type=15
[INFO] Lap 1 completed: 81.234s (P1)
[INFO] Lap 2 completed: 80.891s (P1)

`pitwall replay sessions`

List all recorded sessions.

pitwall replay sessions [OPTIONS]

Option	Default	Description
`-d, --data-dir`	./pitwall_data	Storage directory

`pitwall replay recordings`

List raw .pitraw recordings available for simulation.

pitwall replay recordings [OPTIONS]

Option	Default	Description
`-d, --data-dir`	./pitwall_data	Storage directory

`pitwall replay lap`

Replay a recorded lap frame-by-frame from Parquet data.

pitwall replay lap <SESSION_DIR> --lap <LAP_NUMBER>

Argument	Description
`session_dir`	Path to session directory
`-l, --lap`	Lap number to replay

`pitwall replay export`

Export telemetry to a standalone Parquet file for analytics.

pitwall replay export <SESSION_DIR> [OPTIONS]

Option	Default	Description
`-l, --lap`	(none)	Specific lap to export. Omit for session summary
`-o, --output`	./export.parquet	Output file path

`pitwall sim`

Simulate a raw recording by sending packets over UDP. Used for development testing without the game running.

pitwall sim <RECORDING> [OPTIONS]

Option	Default	Description
`-p, --port`	20777	Target UDP port to send packets to
`-s, --speed`	1.0	Playback speed multiplier

Typical dev workflow (three terminals):

# Terminal 1: receiver in dry-run mode
pitwall live f1 --dry-run

# Terminal 2: simulate recorded packets
pitwall sim ./pitwall_data/recordings/abc-123-def.pitraw

# Terminal 3: live UI
cd pitwall-ui && npm run dev

F1 24 Game Setup

In F1 24, go to Settings → UDP Telemetry:

Setting	Value
UDP Telemetry	On
UDP Broadcast Mode	Off
UDP IP Address	Your computer's IP (e.g. 192.168.1.100)
UDP Port	20777
UDP Send Rate	20Hz (or higher)
UDP Format	2024

For local testing (game on same machine), use 127.0.0.1 as the IP.

Querying with DuckDB

The Parquet files are directly queryable:

-- Load a session
SELECT * FROM 'pitwall_data/sessions/2026/monza/abc-123-def/telemetry.parquet'
WHERE lap_number = 3
ORDER BY timestamp;

-- Best braking point into T1
SELECT lap_distance, speed, brake, g_longitudinal
FROM 'telemetry.parquet'
WHERE lap_number = 3 AND lap_distance BETWEEN 800 AND 900;

-- Tyre temp comparison across laps
SELECT lap_number, AVG(tyre_surface_fl) as avg_fl, AVG(tyre_surface_fr) as avg_fr
FROM 'telemetry.parquet'
GROUP BY lap_number;

-- Lap times
SELECT * FROM 'pitwall_data/sessions/2026/monza/abc-123-def/laps.parquet';

Circuit SVGs

Track outline SVGs are auto-generated from GeoJSON source files during cargo build via build.rs.

pitwall/circuits/
├── geojson/       ← source track outlines (checked in)
│   └── sakhir.geojson
└── svg/           ← generated at build time (gitignored)
    └── sakhir.svg

To add a new track: drop a .geojson file with a LineString geometry into circuits/geojson/ and rebuild. The build script normalizes coordinates to a 0–1 space with aspect-ratio correction and outputs a clean SVG polyline.

Architecture

pitwall/src/
├── main.rs          CLI (clap) - command routing
├── build.rs         Build script - generates circuit SVGs from GeoJSON
├── packets.rs       F1 24 UDP structs (repr(C, packed), zero-copy via bytemuck)
├── parsers.rs       Zero-copy packet parsing from raw bytes
├── session.rs       Session state machine, sample merging, lap detection
├── storage.rs       Parquet + JSON writer (ZSTD compressed, columnar)
├── receiver.rs      Async UDP listener (tokio), drives session lifecycle
├── recorder.rs      Raw packet recording (.pitraw) and UDP simulation
├── replay.rs        Frame-by-frame lap reconstruction from Parquet
├── export.rs        Lap export, session summary, session discovery
└── live_server.rs   WebSocket server (axum) broadcasting LiveFrame JSON

pitwall-ui/src/
├── App.tsx          Main dashboard layout
├── hooks/
│   └── useTelemetry.ts   WebSocket connection + auto-reconnect
└── components/
    ├── SessionHeader.tsx  Track, session type, live indicator, clock
    ├── SpeedPanel.tsx     Speed/RPM/Gear with animated RPM bar
    ├── InputBars.tsx      Throttle/Brake bars
    ├── ErsPanel.tsx       ERS, DRS, fuel status
    ├── TyreTemps.tsx      4-wheel temp grid with heat colors
    └── LapTable.tsx       Live sector timing + purple highlights

License

Personal project. Not affiliated with EA, Codemasters, or Formula 1.

Name		Name	Last commit message	Last commit date
Latest commit History 4 Commits
.kiro/steering		.kiro/steering
pitwall-ui		pitwall-ui
pitwall		pitwall
.gitignore		.gitignore
README.md		README.md

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

PitWall

What It Does

Live UI

Telemetry Channels Captured

Storage Layout

Installation

Commands

`pitwall live f1`

`pitwall replay sessions`

`pitwall replay recordings`

`pitwall replay lap`

`pitwall replay export`

`pitwall sim`

F1 24 Game Setup

Querying with DuckDB

Circuit SVGs

Architecture

License

About

Uh oh!

Releases

Packages

Uh oh!

Contributors

Uh oh!

Languages

Folders and files

Latest commit

History

Repository files navigation

PitWall

What It Does

Live UI

Telemetry Channels Captured

Storage Layout

Installation

Commands

pitwall live f1

pitwall replay sessions

pitwall replay recordings

pitwall replay lap

pitwall replay export

pitwall sim

F1 24 Game Setup

Querying with DuckDB

Circuit SVGs

Architecture

License

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

`pitwall live f1`

`pitwall replay sessions`

`pitwall replay recordings`

`pitwall replay lap`

`pitwall replay export`

`pitwall sim`

Packages