An artificial intelligence (AI)-driven system that promises to transform how factories operate has been developed by a team of researchers at the University of Virginia.

By using Multi-Agent Reinforcement Learning (MARL), the researchers created a way to optimize manufacturing systems, thus improving both speed and quality while simultaneously reducing waste.

To accomplish this, multiple AI agents were coordinated to manage tasks in real time. Because the system can reportedly adjust automatically, learning and improving performance over time, the team suggests that the system could lead to faster production, reduced downtime and better-quality products across many industries.

The researchers explained: "We are addressing the complexity of modern manufacturing. Instead of optimizing individual processes in isolation, our system looks at the big picture — coordinating everything at once. The result is smarter, faster and more adaptable manufacturing."

The algorithms developed — the Credit-Assigned Multi-Agent Actor-Attention-Critic (C-MAAC) and the Physics-Guided Multi-Agent Actor-Attention-Critic (P-MAAC) — reportedly made this possible by allowing the system to account for the physical constraints of machinery as well as the unpredictable production disruptions.

"By integrating system- and process-level parameters, this system can optimize yields and dynamically adapt to changes, such as machine breakdowns or production adjustments, without human intervention. It's a major leap forward in smart manufacturing," the researchers added.

Its developers suggest that in addition to improving productivity, the system also promises to offer economic and environmental advantages. Specifically, by reducing waste, minimizing downtime and lowering energy consumption, manufacturers could potentially achieve cost savings while also reducing their environmental footprint.

An article detailing the technology, “Multi-agent reinforcement learning for integrated manufacturing system-process control,” appears in the Journal of Manufacturing Systems.

To contact the author of this article, email mdonlon@globalspec.com