jevbooks

← All projects

jev-pilot-reflex

manhua-man/jev-pilot-reflex

Three.js autonomous driving reflex and AI safety brake simulator.

Agent Decisions36%Runner-up: Game & Simulation

What it is

An open-source 3D embodied driving simulator for autonomous driving researchers and developers, featuring a Tesla Model Y, Ackermann steering, procedural city streets, and a highway environment.

How it uses Jev

Jev acts as the System 1 local reflex cerebellum: it takes structured telemetry state and returns typed reflex decisions (e.g., ADAPTIVE_CRUISE, EMERGENCY_BRAKE, SWERVE_EVADE) with confidence, latency, throttle, brake, and target lane, used to control the vehicle in the 60 FPS loop.

Technique worth stealing

Dual-brain decoupling: fast local reflex (System 1) with an asynchronous slow strategy brain (System 2) that never blocks the main loop.

Try it

git clone, npm install, npm run dev, open http://localhost:5173

View on GitHub

judged by Jevjev-1.13.0

Evidence

Each line is one question put to Jev about the README. ≥ 0.60 reads as yes, ≤ 0.40 as no; in between Jev is not making a call.

  • Jev-centricyes0.76
  • Shows a System One patternyes0.72
  • Handles uncertaintyno0.08
  • Measuredno0.12
  • Runnableunclear0.43
  • Worth recommendingno0.32
  • Model replicano0.38
  • Problem scopescore on a 0–2 scale1.08
  • About Jevyes0.96

Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 23 Sept 2026.