Data Formats

All data in the TracingInsights repository is stored in JSON format with consistent structures and field naming conventions. This guide covers the format of each data type and how to parse them.

JSON Structure Overview

Every JSON file follows a predictable structure where data is organized in arrays, with each array representing a specific field across all samples.

Array-Based Format

Rather than storing each sample as an object, data is stored in parallel arrays for efficiency:

{
  "tel": {
    "time": [0.0, 0.26, 0.52, 0.78, ...],
    "speed": [0, 42, 89, 134, ...],
    "throttle": [0, 45, 78, 100, ...]
  }
}

This means:

Index 0 across all arrays = first sample
Index 1 across all arrays = second sample
Index n across all arrays = nth sample

Telemetry Data Format

File: `X_tel.json`

Contains second-by-second sensor data for a single lap (~3.7 Hz sampling rate, approximately every 270ms).

Structure
Field Definitions
Example Usage

{
  "tel": {
    "time": [0.0, 0.26, 0.52, ...],
    "rpm": [11245, 11567, 11892, ...],
    "speed": [0, 42, 89, ...],
    "gear": [1, 2, 3, ...],
    "throttle": [0, 45, 78, ...],
    "brake": [false, false, true, ...],
    "drs": [0, 0, 12, ...],
    "distance": [0.0, 12.4, 28.9, ...],
    "rel_distance": [0.0, 0.002, 0.005, ...],
    "x": [-3650.5, -3648.2, -3645.8, ...],
    "y": [1193.7, 1195.3, 1197.1, ...],
    "z": [2.1, 2.3, 2.5, ...],
    "acc_x": [5.2, 4.8, 4.5, ...],
    "acc_y": [0.3, 0.5, 1.2, ...],
    "acc_z": [0.1, 0.2, 0.3, ...],
    "DriverAhead": ["VER", "VER", "VER", ...],
    "DistanceToDriverAhead": [125.4, 123.8, 122.1, ...],
    "dataKey": "2026-Australian Grand Prix-Practice 1-NOR-1"
  }
}

Field	Type	Unit	Description
`time`	float	seconds	Time from start of lap
`rpm`	integer	RPM	Engine revolutions per minute
`speed`	integer	km/h	Car speed
`gear`	integer	1-8	Current gear (F1 uses 8-speed gearboxes)
`throttle`	integer	0-100	Throttle position percentage
`brake`	boolean	-	Brake application (true/false)
`drs`	integer	0-14	DRS status (0,1=Off; 10,12,14=On)
`distance`	float	meters	Distance from lap start
`rel_distance`	float	0.0-1.0	Normalized distance (0=start, 1=end)
`x`	float	meters	Horizontal position (left-right)
`y`	float	meters	Horizontal position (forward-back)
`z`	float	meters	Vertical height/elevation
`acc_x`	float	m/s²	Longitudinal acceleration (forward/backward)
`acc_y`	float	m/s²	Lateral acceleration (side-to-side)
`acc_z`	float	m/s²	Vertical acceleration (up/down)
`DriverAhead`	string/null	-	3-letter code of driver ahead
`DistanceToDriverAhead`	float/null	meters	Gap to car ahead
`dataKey`	string	-	Unique identifier for this lap

import json

# Load telemetry file
with open('NOR/1_tel.json', 'r') as f:
    data = json.load(f)

tel = data['tel']

# Get all speed samples
speeds = tel['speed']

# Get speed at 10th sample
speed_at_10 = tel['speed'][10]

# Find maximum speed on this lap
max_speed = max(tel['speed'])

# Iterate through all samples
for i in range(len(tel['time'])):
    time = tel['time'][i]
    speed = tel['speed'][i]
    throttle = tel['throttle'][i]
    print(f"At {time}s: {speed} km/h, {throttle}% throttle")

Important: Acceleration data (acc_x, acc_y, acc_z) is computed from position and speed data, not raw IMU sensors. Values are smoothed and outliers are filtered.

Data Key Format

The dataKey field uniquely identifies each lap:

"Year-EventName-Session-DriverCode-LapNumber"

Examples:

"2026-Australian Grand Prix-Race-VER-9"
"2026-PreSeasonTesting1-Practice 1-NOR-5"

Lap Times Data Format

File: `laptimes.json`

Contains comprehensive timing and performance data for all laps completed by a driver.

Structure
Core Timing Fields
Tire Fields
Position & Status Fields

{
  "time": [101.411, 84.906, 84.391, ...],
  "lap": [1, 2, 3, ...],
  "compound": ["MEDIUM", "MEDIUM", "SOFT", ...],
  "stint": [1, 1, 2, ...],
  "s1": [29.609, 29.862, 29.637, ...],
  "s2": [18.321, 18.361, 17.959, ...],
  "s3": [36.976, 36.168, 35.795, ...],
  "life": [5, 6, 1, ...],
  "pos": [1, 1, 2, ...],
  "status": ["1", "1", "2", ...],
  "pb": [true, false, false, ...],
  "fresh": [false, false, true, ...],
  "drv": ["NOR", "NOR", "NOR", ...],
  "dNum": ["1", "1", "1", ...],
  "team": ["McLaren", "McLaren", "McLaren", ...]
}

Field	Type	Unit	Description
`time`	float/string	seconds	Total lap time (“None” for incomplete laps)
`lap`	integer	-	Lap number
`sesT`	float	seconds	Session time when lap ended
`lST`	float	seconds	Session time when lap started
`lSD`	string	ISO datetime	Calendar timestamp of lap start
`s1`	float/string	seconds	Sector 1 time
`s2`	float/string	seconds	Sector 2 time
`s3`	float/string	seconds	Sector 3 time
`s1T`	float	seconds	Session time when S1 was set
`s2T`	float	seconds	Session time when S2 was set
`s3T`	float	seconds	Session time when S3 was set

Field	Type	Description
`compound`	string	”SOFT”, “MEDIUM”, “HARD”, “INTERMEDIATE”, “WET”, “UNKNOWN”
`life`	integer	Total laps driven on this tire set
`fresh`	boolean	Was this a new tire when fitted?
`stint`	integer	Stint number (increments after each pit stop)

Field	Type	Description
`pos`	integer/string	Track position at lap end (“None” for practice/quali)
`status`	string	Track status codes during this lap (e.g., “1”, “2”, “126”)
`pb`	boolean	Is this the driver’s official personal best lap?
`del`	boolean	Was this lap deleted by stewards?
`delR`	string/null	Reason for deletion if applicable
`iacc`	boolean	Is timing data considered accurate?
`ff1G`	boolean	Was this lap generated/interpolated by FastF1?

Track Status Codes

Code	Meaning	Description
`"1"`	Track Clear	Normal racing conditions
`"2"`	Yellow Flag	Hazard on track, no overtaking
`"4"`	Safety Car	Physical safety car deployed
`"5"`	Red Flag	Session stopped
`"6"`	VSC Deployed	Virtual Safety Car active
`"7"`	VSC Ending	VSC about to end

Multiple status codes may appear together (e.g., "126" = Track Clear + Yellow + VSC). Each digit represents a different status.

Weather Data Format

File: `weather.json`

Environmental conditions sampled approximately once per minute.

{
  "wT": [3.099, 63.102, 123.101, ...],
  "wAT": [20.1, 20.2, 20.2, ...],
  "wH": [73.2, 72.9, 72.3, ...],
  "wP": [1015.0, 1015.1, 1015.0, ...],
  "wR": [false, false, false, ...],
  "wTT": [25.3, 25.8, 26.1, ...],
  "wWD": [180, 185, 190, ...],
  "wWS": [2.5, 2.8, 3.1, ...]
}

Field	Full Name	Unit	Description
`wT`	Time	seconds	Session time when sampled
`wAT`	Air Temperature	°C	Ambient air temperature
`wH`	Humidity	%	Relative humidity
`wP`	Pressure	mbar	Atmospheric pressure
`wR`	Rainfall	boolean	Is it raining?
`wTT`	Track Temperature	°C	Asphalt surface temperature
`wWD`	Wind Direction	degrees	Compass bearing (0-359)
`wWS`	Wind Speed	m/s	Wind velocity

Race Control Messages Format

File: `rcm.json`

Official messages from race control about flags, incidents, and track status.

{
  "time": ["2026-03-06T01:15:21.000000000", ...],
  "cat": ["Flag", "Other", "SafetyCar", ...],
  "msg": ["GREEN", "DRS ENABLED", ...],
  "status": ["ENABLED", "DISABLED", ...],
  "flag": ["GREEN", "YELLOW", "RED", ...],
  "scope": ["Track", "Sector", "Driver", ...],
  "sector": [null, 2, null, ...],
  "dNum": [null, null, "1", ...],
  "lap": [null, 5, null, ...]
}

Field	Type	Description
`time`	ISO datetime	When message was issued
`cat`	string	”Flag”, “Other”, “SafetyCar”, “Drs”, “CarEvent”
`msg`	string	Human-readable message text
`status`	string	Additional context (“ENABLED”, “DISABLED”, etc.)
`flag`	string	Flag type if applicable
`scope`	string	”Track”, “Sector”, or “Driver”
`sector`	integer/null	Mini-sector number (1-24) if scope is “Sector”
`dNum`	string/null	Driver number if scope is “Driver”
`lap`	integer/null	Lap number if applicable

Corners Data Format

File: `corners.json`

Circuit layout information with corner positions and angles.

{
  "CornerNumber": [1, 2, 3, 4, 5, ...],
  "X": [-3650.59, -3555.05, -7194.97, ...],
  "Y": [1193.78, 1992.75, 7101.94, ...],
  "Angle": [-176.63, -1.52, 162.26, ...],
  "Distance": [350.25, 424.90, 1067.86, ...],
  "Rotation": 44.0
}

Field	Type	Unit	Description
`CornerNumber`	integer	-	Sequential corner identifier (1, 2, 3…)
`X`	float	meters	Horizontal position on track map
`Y`	float	meters	Horizontal position on track map
`Angle`	float	degrees	Visual offset angle for map markers
`Distance`	float	meters	Distance from start/finish line
`Rotation`	float/null	degrees	Track rotation to match official map

Drivers Data Format

File: `drivers.json`

Metadata about all drivers participating in the session.

{
  "drivers": [
    {
      "driver": "NOR",
      "team": "McLaren",
      "dn": "1",
      "fn": "Lando",
      "ln": "Norris",
      "tc": "F47600",
      "url": "https://media.formula1.com/.../lannor01.png"
    },
    ...
  ]
}

Field	Full Name	Type	Description
`driver`	Driver Code	string	3-letter identifier (“NOR”, “VER”, etc.)
`team`	Team Name	string	Constructor name
`dn`	Driver Number	string	Car number
`fn`	First Name	string	Driver’s first name
`ln`	Last Name	string	Driver’s last name
`tc`	Team Color	string	Hex color code (without #)
`url`	Headshot URL	string	Link to driver photo

Unlike other files, drivers.json uses an object-based format with an array of driver objects rather than parallel arrays.

Field Naming Conventions

The data uses abbreviations consistently:

Prefixes:
- w = Weather (e.g., wT, wAT, wH)
- s = Sector (e.g., s1, s2, s3)
- v = Speed/Velocity (e.g., vi1, vfl, vst)
- p = Pit (e.g., pin, pout)
- l = Lap (e.g., lST, lSD)
- d = Driver (e.g., dNum, drv)
Suffixes:
- T = Time/Temperature (context-dependent)
- D = Date/Direction/Distance (context-dependent)
- S = Speed/Start (context-dependent)

Parsing Tips

Handle Missing Values

Many fields can be null or "None" (as a string):

# Check for missing lap time
if lap_time != "None" and lap_time is not None:
    print(f"Lap time: {lap_time}s")

Align Array Indices

When working with telemetry or laptimes, remember all arrays share the same indices:

for i in range(len(tel['time'])):
    sample = {
        'time': tel['time'][i],
        'speed': tel['speed'][i],
        'throttle': tel['throttle'][i],
        'brake': tel['brake'][i]
    }
    process_sample(sample)

Convert Data Types

JSON doesn’t distinguish between int and float - handle accordingly:

import json

data = json.load(f)

# Ensure integers where needed
lap_number = int(data['lap'][0])

# Handle boolean strings
is_raining = data['wR'][0]  # Already boolean in JSON

Accessing Data

Analysis

JSON Structure Overview

Array-Based Format

Telemetry Data Format

File: `X_tel.json`

Data Key Format

Lap Times Data Format

File: `laptimes.json`

Track Status Codes

Weather Data Format

File: `weather.json`

Race Control Messages Format

File: `rcm.json`

Corners Data Format

File: `corners.json`

Drivers Data Format

File: `drivers.json`

Field Naming Conventions

Parsing Tips

Handle Missing Values

Align Array Indices

Convert Data Types

Next Steps

Field Reference

Python Examples

Build docs developers (and LLMs) love

Accessing Data

Analysis

​JSON Structure Overview

​Array-Based Format

​Telemetry Data Format

​File: X_tel.json

​Data Key Format

​Lap Times Data Format

​File: laptimes.json

​Track Status Codes

​Weather Data Format

​File: weather.json

​Race Control Messages Format

​File: rcm.json

​Corners Data Format

​File: corners.json

​Drivers Data Format

​File: drivers.json

​Field Naming Conventions

​Parsing Tips

​Handle Missing Values

​Align Array Indices

​Convert Data Types

​Next Steps

Field Reference

Python Examples

Build docs developers (and LLMs) love

JSON Structure Overview

Array-Based Format

Telemetry Data Format

File: `X_tel.json`

Data Key Format

Lap Times Data Format

File: `laptimes.json`

Track Status Codes

Weather Data Format

File: `weather.json`

Race Control Messages Format

File: `rcm.json`

Corners Data Format

File: `corners.json`

Drivers Data Format

File: `drivers.json`

Field Naming Conventions

Parsing Tips

Handle Missing Values

Align Array Indices

Convert Data Types

Next Steps