SmartPilot:Agent-Based CoPilot for Intelligent Manufacturing

Chathurangi Shyalika (Artificial Intelligence Institute, University of South Carolina), Renjith Prasad (Artificial Intelligence Institute, University of South Carolina), Alaa Al Ghazo (Artificial Intelligence Institute, University of South Carolina), Darssan L. Eswaramoorthi (Artificial Intelligence Institute, University of South Carolina), Sara Shree Muthuselvam (Artificial Intelligence Institute, University of South Carolina), Amit Sheth (Artificial Intelligence Institute, University of South Carolina)

Abstract

In the dynamic landscape of Industry 4.0, achieving efficiency, precision, and adaptability is essential for optimizing manufacturing operations. SmartPilot is a neurosymbolic and agent-based CoPilot designed to enhance real-time decision-making capabilities in manufacturing. The system addresses three key challenges: anomaly prediction, production forecasting, and domain-specific question answering through an agent-based framework. SmartPilot leverages multimodal data and a compact architecture optimized for edge devices. This paper highlights its innovative combination of agent-based design and neurosymbolic reasoning to enable contextual decision-making in complex environments. The demonstration video 1 , datasets, and supplementary materials are available at https://github.com/ChathurangiShyalika/SmartPilot.