1 Native SDK Positioning
The "internal nervous system" of the robot, allowing algorithm engineers to dive deep into the framework and deploy high-performance control and RL policies.
Target users: robot algorithm/control engineers, RL policy developers, and those who need deeply customized control logic.
Typical scenarios: deploying RL walking/dancing policies, custom whole-body controllers, algorithm validation, and real-robot migration.
2 Open-Source Repository and Module Introduction
This repository is the EngineAI native robot control SDK, designed for humanoid robot application development and system integration. It provides a lightweight, easy-to-deploy, and highly extensible control and task execution framework. Through standardized interfaces and a modular architecture design, the SDK effectively reduces the complexity of secondary development and functional integration of robot systems, allowing developers to focus more on algorithm research and application feature implementation.
Core Modules:
- High-performance precompiled scheduling framework Provides a high-performance scheduling mechanism for real-time robot control and inference tasks, reducing runtime overhead through precompilation optimization, and improving overall system execution efficiency and stability.
- Configuration-driven task orchestration system Adopts a configuration-based task orchestration approach to achieve flexible organization of algorithm modules, control flows, and data streams, supporting rapid adjustment of experiment flows and system runtime strategies.
- Extensible model and parameter management system Provides unified interfaces for model loading, version management, and parameter configuration, facilitating centralized management and iteration of algorithm models, control parameters, and experiment configurations.
- Modular business plugin mechanism Based on a plugin-based architecture design, it supports flexible extension and decoupling of functional modules such as perception, planning, and control, facilitating rapid integration of new algorithms or new features.
- System data interface and runtime monitoring. Supports standardized data interaction with external systems and provides runtime state observation and visual debugging capabilities. It can be combined with the ROS2 ecosystem for module communication, and supports real-time monitoring and analysis of key states, control quantities, and feedback data via PlotJuggler, improving joint-debugging efficiency and problem localization capabilities.
- Mujoco simulation and real-robot deployment toolchain Provides a complete Mujoco simulation environment and real-robot deployment scripts, supporting rapid migration from simulation verification to real robot systems; in addition, the repository also contains robot model files such as URDF that are consistent with the real robot hardware structure, used for simulation environment construction, kinematics/dynamics computation, and algorithm verification, ensuring good consistency between simulation results and the real system.