NEUP CINR-20-19837: Reinforcement Learning Validation Framework for Quality Assurance of AI-guided Additive Manufacturing Digital Platforms

This project proposes a mathematically-rigorous validation framework -- based on a multi-armed bandit reinforcement learning strategy -- which assimilates all sensors data and analysis results obtained from the digital platform to calculate risk measures in the form of failure probabilities for the AM components. The project will develop and demonstrate the proposed reinforcement learning framework using TCR data and its associated digital platform which will be installed at Purdue campus.

Additional Info

Field	Value
Awarded Institution	Purdue University
PI	John Sutherland