Home

Welcome to the MoSy Engine Wiki

MoSy (Motion Sync Edge AI Engine) is a high-performance, low-latency control platform designed to execute Reinforcement Learning (RL) quadruped locomotion policies on resource-constrained edge hardware.

Core Objectives

• Sub-millisecond Inference: Execute pre-trained RL motor policies via ONNX Runtime without heavy deep learning frameworks.
• Hardware Integration: Seamlessly map wearable motion-tracking telemetry (HS-1 suit featuring BNO055 IMUs) to quadruped state-space vectors.
• Cross-Platform Readiness: Maintain decoupled Python and C++ inference pipelines for desktop simulation and embedded deployment.

---

Documentation Index

1. System Architecture — Deep dive into the 112-to-8 tensor pipeline.
2. Getting Started — Environment setup and model conversion guide.
3. Telemetry Mapping — BNO055 sensor fusion and observation vector layout.
4. Specification — Orchestrates the system pipeline and enforces a strict.