Back to the archive

The Telemetry Circuit / Working prototype

PPO racing agent driven by live telemetry

An autonomous F1 22 racing agent that reads live UDP telemetry and drives through a virtual Xbox controller.

PythonStable Baselines3GymnasiumUDP Telemetryvgamepad
* CHECK
SUBJECT
F1 Agent
TYPE
Applied reinforcement learning
STATUS
Working prototype

* 60 Hz telemetry processing.

PROOF SIGNAL

60 Hztelemetry processing

The mission

What the system does

An autonomous F1 22 racing agent that reads live UDP telemetry and drives through a virtual Xbox controller.

  • 60 Hz game telemetry parser
  • Custom Gymnasium environment
  • Virtual-LiDAR track-deviation features
  • Continuous steering, throttle, and braking

Engineering log

Decisions behind the build

What the system handles, and why it works this way.

BOSS 01

Seeing the track

Raw game packets are not a useful driving representation. Virtual-LiDAR features turn track deviation into an observation the policy can learn from.

BOSS 02

Real-time control

Telemetry and controller output must remain synchronized. A steady loop prevents stale observations from producing unstable actions.

System map

How signals move

A plain technical view. The map still works without hovering.

01
F1 22UDP

Streams live telemetry

02
ParserPython

Builds 60 Hz state

03
Track modelVirtual LiDAR

Represents boundaries

04
PolicyPPO

Chooses continuous controls

05
Drivervgamepad

Sends controller input

LOG ENTRY / WHAT I LEARNED

“A strong observation space can matter more than a more complicated learning algorithm.”

Source signal

Inspect the work.