Human-Machine-AI Collaborative Manufacturing
Immersive & Interactive Cyber-Physical System (I2CPS)
- Virtual Reality (VR) interface for robotic manufacturing [Paper]
- To visualize machine data and environments for intuitive decision making
- To transfer virtual works to robot’s automation, enhancing flexibility
- To digitalize human skills and develop AI
- Application #1: Exo-skeleton for wider working range of humans [Paper]
- Application #2: Robotic finishing to transferring virtual manual works into a robot’s automation
Autonomous Robotic Bin Picking
- Finding the most pickable object with machine vision and machine learning (ML)
- Solutions for flexible machine tending in automation devices (e.g. machine tools)
- Human demonstration for fast training of ML [Paper #1] [Paper #2]
Digital Twin of Robotic Manufacturing
Virtual commissioning
- Running robot script virtually
- Collision prediction from given robot script
- Toward planning of collaboration with multiple manipulators via ML
Generation of synthetic 2D/3D images
- Building virtual environment with high speed and accuracy
- Pre-training AI with synthetic data [Challenges]
Manufacturing AI
Condition-based Monitoring (CBM)
- Introducing new data processing and ML technologies to CBM
- Understanding complex manufacturing processes [Paper #1] [Paper #2]
- In-situ prediction of productivity [Paper] and anomalies [Paper]
- Tool Condition Monitoring (TCM) [Challenges]
Visual Quality Inspection
- Combination of image processing and ML
- Synthetic data to overcome limited amount of data and unbalances [Paper #1] [Paper #2]
- Auto labeling of training data [Paper]
- Explainable AI for reliabile usage of ML in industry [Paper]
Investigating Manufacturability
- ML-based Computer-Aided Process Planning (CAPP) [Paper]
- Assessment of manufacturability from generative design
