Generative AI Models for On-Demand Design in Lights-Out Manufacturing Facilities
The rise of lights-out manufacturing—facilities operating autonomously without human intervention—has redefined the landscape of industrial automation. However, these systems often rely on pre-designed parts and rigid workflows, limiting their adaptability. This paper explores the integration of generative AI models into lights-out manufacturing environments to enable on-demand, real-time product design. By leveraging the capabilities of neural networks such as Generative Adversarial Networks (GANs) and diffusion-based models, manufacturing systems can autonomously create, evaluate, and iterate product designs without human input. We propose an architectural framework for integrating AI-driven design generation with digital twin-based production pipelines, highlight key use cases such as rapid prototyping and design optimization, and assess the technical challenges associated with quality control, validation, and data integrity. Our analysis suggests that generative AI can dramatically enhance the responsiveness and efficiency of fully automated facilities, marking a significant step toward fully autonomous product lifecycles.