仿生机器人配件数据库

language: - en - zh license: cc-by-4.0 library_name: datasets tags: - robotics - bionic-robot - actuators - sensors - embedded-systems - llm - robot-components - embodied-ai - ros2 - hardware-selection - lerobot - robot-ai-models - flexible-actuators - humanoid-robot datasets_info: - config_name: default features: - name: category dtype: string - name: entities sequence: - name: name dtype: string - name: manufacturer dtype: string - name: type dtype: string - name: specs dtype: struct - name: source dtype: string - name: confidence dtype: float32 - name: last_verified dtype: string - name: standard_compliance sequence: string - name: bionic_features sequence: string - name: ros2_compatible dtype: bool splits: - name: full num_bytes: 380000 num_examples: 688 --- # RoboParts Dataset **RoboParts** is the first structured dataset covering ten core categories of bionic robot components: actuators, flexible actuators, sensors, chips, communication protocols, large language models (LLMs), robot AI models, interfaces, platforms, and data acquisition devices. Designed for embodied intelligence research and humanoid robot development. ## Dataset Overview | Category | Count | Key Fields | |---|---|---| | Actuators | 199 | Torque, speed, voltage, protocol, weight, price | | Flexible Actuators | 21 | Flexibility, deformation, drive type, material | | Sensors | 90 | Range, precision, type, manufacturer | | Chips | 108 | CPU, AI performance, power, price range | | Protocols | 64 | Speed, latency, max nodes, standard | | LLMs | 42 | Parameters, open-source, embodied support | | Robot AI Models | 44 | Model type, parameters, tasks, open-source | | Interfaces | 37 | Speed, power, connector type | | Platforms | 40 | Type, open-source, simulation support | | Data Acquisition | 43 | Teleoperation, motion capture, tactile sensing | **Total: 688 entities across 10 categories** ## LeRobot Ecosystem This dataset is tagged for the [LeRobot](https://github.com/huggingface/lerobot) robotics learning ecosystem (v0.6.0+). Compatible with world model policies (VLA-JEPA, LingBot-VA, FastWAM) and reward model APIs introduced in v0.6.0. Use `tags: lerobot, robotics` to discover related models and datasets on HuggingFace. ## Data Quality - **Standardized fields**: Every entity includes `source`, `confidence`, `last_verified`, `standard_compliance` - **Bionic tagging**: Supports filtering for bionic actuators (SEA / flexible) - **National standard compliance**: Tracks GB humanoid robot modular requirements - **ROS2 compatibility**: Annotated for ROS2 compatibility - **Continuous updates**: Collected from manufacturer datasheets, ROS2 docs, ISO/national standards | Field | Coverage | |---|---| | Name | 100% | | Manufacturer | 95% | | Specifications | 85% | | Price range | 70% | | Standard compliance | 60% | ## Usage ### Direct Download ```python from datasets import load_dataset dataset = load_dataset("roboparts/roboparts-dataset", split="full") ``` ### Python (Raw JSON) ```python import json, requests url = "https://roboparts.cc/api/data.json" data = requests.get(url).json() # Find all bionic actuators bionic_actuators = [ a for a in data["actuators"] if a.get("bionic_features") and len(a["bionic_features"]) > 0 ] ``` ## Smart Selection Engine The dataset powers the online selection engine at [roboparts.cc](https://roboparts.cc): - Input joint position, torque requirement, budget - Automatic compatible actuator matching - Multi-factor scoring (torque/price 30%, weight 20%, protocol match 25%, bionic 15%, standard 10%) ## Citation ```bibtex @dataset{roboparts2026, title={RoboParts: A Structured Dataset for Bionic Robot Components}, author={RoboParts Team}, year={2026}, url={https://roboparts.cc} } ``` ## License [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) — Free to use with attribution. ## Links - Website: https://roboparts.cc - GitHub: https://github.com/lm203688/roboparts - ModelScope (public): https://www.modelscope.cn/datasets/roboparts/roboparts-data - API Docs: https://roboparts.cc/api-pricing