Technology Overview
The Intelligent Scheduling System includes a learning mechanism that enables an optimal allocation of factory machines and production resources. Two major algorithms are employed: (1) Monte Carlo tree search (MCTS) algorithm is used to search best results via playout and construct a daily master production plan; and (2) Deep Neural Networks (DNN) regression models are built to explore implicit production Work In Process (WIP) information and determine dispatch strategy. They can steadily improve production capacity and on-time delivery and contribute to net zero carbon emissions with better production process.
Applications & Benefits
This system has been applied to industries such as display panel, steel, pharmaceutical, and medical devices for verification.