Human-AI Collaboration on Leveraging Human Operators’ Tacit Knowledge During Machine Adjustment in Manufacturing
Machine adjustment process (Setup process) in some manufacturing business requires expertised tacit knowledge of human operators.
Due to the characteristic of tacit knowledge which is hard to be expressed in words, it’s challenging to manualize the process or transfer their self-developed strategy to entry-level workers.
To address this, our work aims to leverage reinforcement learning algorithms to imitate human expert’s work and make the algorithm find the optimal policy without hard-coding the rules.
When the algorithm is set, we further aim to make it more easily communicable with both expert and entry-level operators, enabling expert operators to ensure stable performance even in difficult scenarios, and allowing entry-level workers to communicate as they are collaborating with a real-time personal instructor.
Related Publications
Researchers
