Data sets from CPS for ML applications
TuLAUT (Theory and Teaching of Automation Technology) is an initiative of university professors in Germany who research and teach in the field of automation technology. What they have in common is that they consider complex technical processes and plants from the point of view of how automation solutions can be designed and implemented for them using suitable methods and which descriptive means and tools can support these methods.
The common goal is to develop methods that can be applied to real, complex, often not formally described engineering tasks of automation engineering practice, be it for automation in machines, process or production engineering plants, in energy generation and distribution, in buildings, vehicles or traffic systems.
For more information about TuLAUT have a look at the TuLAUT Website
Here on this page we have a curated list of data sets which have been used and/or generated by members of the TuLAUT group. We distinguish the data by sets by task (anomaly detection (supervised & unsupervised), diagnosis, planning, product monitoring, production optimization). Have a look on the individual page of a data set for more information about it and how to use it.
| Data Set | Task | Domain |
|---|---|---|
| wdn-sa-benchmark | diagnosis, planning, reconfiguration | process plants |
| HAI-CPS | diagnosis, planning, reconfiguration | process plants |
| IndPenSim | anomaly detection, diagnosis | chemical plant |
| Pick and Place Unit | - | manufacturing engineering (demonstrator) |
| Metal Classification | classification | material science |
| Smart Elevator | prediction | building automation |
| Smart Factory | anomaly detection | process plant |
| PCB Production Prediction | prediction | process plant |
| Electric Current Prediction | prediction | production |
| Robotic Wrapping | partitioning, clustering | production configuration |
| Stuttgart Open Relay Degradation Dataset | condition monitoring, transfer learning | electromechanical relays |
| Benchmark Model to Generate Batch Process Data | - | Batch Process |
| Prereformer Reactor Surrogate Model | - | prereformer reactor |
| P&ID | - | distillation plant |
| DN15 extraction column images | - | chemistry |
| Flow Regime Classification | - | bioreactor |
| Vega shrink-wrapper component degradation | system degradation | production plant |
| Versatile Production System | model learning, anomaly detection | production plant |
| Broken bar in a three-phase induction motor | anomaly detection | CPS |
| High Storage System Data for Energy Optimization | anomaly detection | production plant |
| Ion Mill Etch Tool | anomaly detection | production plant |
| Cobot Anomaly | anomaly detection, diagnosis | colaborative robot |
| Proces Engineering Benchmarks | fault handling | process engineering |
| Building Fault Detection and Diagnostics | diagnosis | building ventilation |
| LiU-ICE | diagnosis | car engine |
| SLIDe | diagnosis | steam line |
| LUMEN | diagnosis | rocket engine |