THULab/spa_francorchamps_lap_data

Spa-Francorchamps Lap Data (TsFile)

Apache TsFile version of Nasim435/spa-francorchamps-lap-data.

Overview

High-frequency sim racing telemetry recorded from Assetto Corsa using shared memory extraction. The dataset contains detailed vehicle dynamics, tire telemetry, suspension behavior, braking data, and spatial racing telemetry collected during 5 laps at Spa-Francorchamps with the Chevrolet Corvette C7.R.

• Rows: 51,554 telemetry samples (one continuous 100 Hz stream across the 5 recorded laps).
• Sampling rate: 100 Hz.
• Recorded laps: 5.
• Telemetry source: Assetto Corsa shared memory via AC-Datalogger.

Track

Property	Value
Track	Spa-Francorchamps
Layout	spa
Recorded laps	5

Vehicle

Property	Value
Car	Chevrolet Corvette C7.R
Brand	Chevrolet
Class	Race / GTE
Drivetrain	RWD
Transmission	Sequential
Max RPM	6500
Power	495+ bhp
Torque	650+ Nm
Weight	1320 kg
Top speed	280+ km/h
Fuel capacity	110 L
Power-to-weight	2.66 kg/hp

Telemetry features

• Driver inputs: throttle, brake, clutch, steer_angle
• Vehicle state: speed_kmh, rpms, gear, fuel
• Vehicle dynamics: g_lat, g_lon, yaw_rate, slip_angle
• Tire telemetry: tyre_core_*, slip_*, psi_*, wear_*
• Suspension: susp_*, ride_height_f, ride_height_r
• Position & motion: pos_x, pos_z, vel_*, lap_progress

Schema (TsFile structure)

• Time (INT64, milliseconds) — the source timestamp, a float64 UNIX epoch in seconds, multiplied by 1000 (epoch_s → ms). It runs continuously across the 5 laps.
• No TAG column. The source has no explicit lap identifier; lap_progress is a per-lap normalized progress fraction rather than a stable device key, so the telemetry is kept as one table / one continuous series.
• All 83 remaining columns (FIELD, DOUBLE or INT64) — speed, RPM, gear, pedal inputs, g-forces, suspension, tire temperatures/pressures/wear, wheel speeds, camber, ride height, fuel, GPS/position, and the rest of the 84-column telemetry schema listed above.

telemetry.csv in the source repository is a duplicate of telemetry.parquet and is not included; only the parquet content was converted.

Usage

Install the Apache TsFile Python SDK (pip install tsfile) and read a converted file:

from pathlib import Path
from tsfile import TsFileReader

path = Path("spa_francorchamps_lap_data.tsfile")
with TsFileReader(str(path)) as reader:
    schemas = reader.get_all_table_schemas()
    print("tables:", list(schemas))
    table_name = next(iter(schemas))
    table = schemas[table_name]
    columns = [column.get_column_name() for column in table.get_columns()]
    print("columns:", columns)
    field_names = [
        column.get_column_name()
        for column in table.get_columns()
        if column.get_column_name() not in {"Time", "time"}
    ]
    if field_names:
        with reader.query_table(table_name, field_names[:3], batch_size=1024) as result:
            batch = result.read_arrow_batch()
            if batch is not None:
                print(batch.to_pandas().head())

Source & license

• Original dataset: https://huggingface.co/datasets/Nasim435/spa-francorchamps-lap-data
• Author / publisher: Nasim435 (dataset publisher)
• License: mit