Forgis2026per-subset
FactoryNet
A multi-embodiment industrial time-series corpus with 56,591 end-to-end task executions (14,847 real, 41,744 simulated) across 7 embodiments, 4 tasks, and 27 annotated anomaly types, unified under a control-theoretic Setpoint-Effort-Feedback-Context (S-E-F-C) schema.
Downloads1K
Episodes56,591
Likes9
Why This Matters for Physical AI
FactoryNet enables cross-embodiment industrial anomaly detection and predictive maintenance through a unified control-theoretic schema that separates commanded intent from realized dynamics across heterogeneous manufacturing systems.
Technical Profile
- Modalities
- proprioceptionforce_torquetime-series
- Robot Embodiments
- UR3UR3eUR5UR10UR30KUKA KR10CNC gantryYu-Cobot
- Action Space
- joint_positions
- Environment
- labindustrialsimulation
- Task Types
- pick_and_placescrewingpeg_insertionmachining
- Episodes
- 56,591
- Data Format
- Parquet
- Annotation Types
- anomaly_labelsfault_labelshealthy_baselinescounterfactual_pairsepisode_metadata
- License
- per-subset
Community Signals
Top 50% by downloads
Access
Need custom proprioception data?
Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.
Request a Sample Pack