RoboCute

Welcome to RoboCute

Make Robotics Cute!

A Python-first 3D AIGC/Robotics development tool with node-based workflow

Python License Status

What is RoboCute? / 什么是 RoboCute?

English:

RoboCute is a Python-first 3D AIGC (AI-Generated Content) and Robotics development tool that adopts a ComfyUI-like node-based server-client architecture. It comes with a self-developed cross-platform graphics engine, runtime, and editor, allowing you to write node algorithms directly in Python, seamlessly integrating with the entire Python ecosystem. The desktop runtime overcomes the inherent limitations of web platforms regarding files and 3D content, providing more convenient interactions and better performance.

中文:

RoboCute 是一个 Python-first 的 3D AIGC（AI生成内容）和机器人开发工具，采用类似 ComfyUI 的节点式 server-client 架构。它配备了自研的跨平台图形引擎、运行时和编辑器，允许您直接用 Python 编写节点算法，无缝接入整个 Python 生态。桌面端运行时摆脱了 Web 平台对文件和 3D 内容的天然局限，提供更便利的交互和更好的性能。

Key Features / 核心特性

English:

中文:

🚀 Quick Start / 快速开始

Installation / 安装

WIP (当前还不能使用)

pip install robocute

Basic Usage / 基本使用

运行一个服务器

import robocute as rbc

mesh_path = "D:/ws/data/assets/models/bunny.obj"
# Create a scene
scene = rbc.Scene()
scene.start()

# Create an entity
mesh_id = scene.load_mesh(mesh_path, priority=rbc.LoadPriority.High)
robot = scene.create_entity("Robot")
# Add tranform component to define its position
scene.add_component(
    robot.id,
    "transform",
    rbc.TransformComponent(
        position=[0.0, 0.0, 0.0],
        rotation=[0.0, 0.0, 0.0, 1.0],
        scale=[1.0, 1.0, 1.0],
    ),
)
# Add render component with mesh reference
render_comp = rbc.RenderComponent(
    mesh_id=mesh_id,
    material_ids=[],
)
scene.add_component(robot1.id, "render", render_comp)

# Start server with editor service
server = rbc.Server(title="RoboCute Server", version="0.2.0")
editor_service = rbc.EditorService(scene)
server.register_service(editor_service)
server.start(port=5555)

# Now you can connect with the C++ editor
print("Server started on port 5555")
print("Start the editor to connect and visualize the scene")

打开rbc_editor，查看场景并连接节点，执行后在Editor上查看结果

rbc_editor

Example: Creating Animation Nodes / 示例：创建动画节点

Headless模式下，可以完全不使用Editor来定义节点连接，这一部分同样也可以通过预先保存的节点图进行执行。方便在远程linux服务器上进行大批量的仿真/AI计算。

import robocute as rbc
import custom_nodes.animation_nodes as animation_nodes

# Build a rotation animation graph
graph_def = rbc.GraphDefinition(
    nodes=[
        rbc.NodeDefinition(
            node_id="entity_input",
            node_type="entity_input",
            inputs={"entity_id": robot.id},
        ),
        rbc.NodeDefinition(
            node_id="rotation_anim",
            node_type="rotation_animation",
            inputs={
                "radius": 2.0,
                "angular_velocity": 1.0,
                "duration_frames": 120,
                "fps": 30.0,
            },
        ),
        rbc.NodeDefinition(
            node_id="anim_output",
            node_type="animation_output",
            inputs={"name": "rotation_test", "fps": 30.0},
        ),
    ],
    connections=[
        rbc.NodeConnection(
            from_node="entity_input",
            from_output="entity",
            to_node="rotation_anim",
            to_input="entity",
        ),
        rbc.NodeConnection(
            from_node="rotation_anim",
            from_output="animation",
            to_node="anim_output",
            to_input="animation",
        ),
    ],
)

# Execute the graph
scene_context = rbc.SceneContext(scene)
graph = rbc.NodeGraph.from_definition(graph_def, "test_graph", scene_context)
executor = rbc.GraphExecutor(graph, scene_context)
result = executor.execute()

Documentation / 文档

Project Status / 项目状态

RoboCute is currently in early development. See Development Log for progress.

RoboCute 目前处于早期开发阶段。查看开发日志了解进度。