Real-Time Revolution: How AI is Mastering Continuous Flow Manufacturing

In this session, Joseph Hernandez and Cissy Gu of Liveline Technologies will share critical insights on applying advanced machine learning to solve complex challenges in real-world manufacturing. They will introduce a proven strategy that transforms existing production assets into autonomous systems, moving beyond traditional methods to a novel, two-stage approach.

First, they will detail how an AI-based model learns the underlying physics of a production line directly from equipment sensors, creating a dynamic "digital twin." Second, they will explain how this digital twin is used to train and validate powerful controllers with reinforcement learning (RL). These autonomous AI agents are then deployed to the factory floor, operating in a closed-loop system to make continuous, real-time adjustments. This lightweight, software-based solution bridges the gap between manual operation and true automation, enabling rapid deployment and flexible updates. Once active, these AI controllers drive production with minimal intervention, constantly compensating for variations in raw materials and processing conditions.

The result is a significant reduction in scrap, waste, and unplanned downtime, empowering the existing plant workforce to manage a broader and more strategic scope of operations.