Deep-Learning-Based Resource Orchestration for Healthcare-Oriented Industrial IoT Flexible Manufacturing Systems With Edge Computing

Rising trends in Industrial Internet-of-things (IIoT), flexible manufacturing, and edge computing present unprecedented opportunities for advancing healthcare manufacturing systems. However, orchestrating resources efficiently while maintaining stringent healthcare quality standards poses significant challenges, particularly in environments where manufacturing processes must adapt quickly to varying production requirements. These challenges are further complicated by the limited computational resources of IIoT devices and the need for real-time processing in medical device production and pharmaceutical manufacturing. To address these challenges, we propose a distributed healthcare-aware deep learning resource orchestration (DH-DLRO) algorithm for edge computing-enabled healthcare IIoT flexible manufacturing systems. Our approach constructs a joint optimization problem for task offloading decisions and resource allocation, specifically tailored to healthcare manufacturing requirements. The algorithm employs multiple parallel deep neural networks to generate efficient offloading decisions while considering healthcare-specific quality constraints and manufacturing precision requirements. The algorithm reduces energy consumption by 25-30% while maintaining medical device manufacturing precision standards, shows remarkable stability across varying task sizes (7000-25000 bytes), and exhibits robust performance under different system parameter configurations. DH-DLRO maintains consistent quality of service levels above 0.95 for medical device assembly tasks while achieving optimal CPU utilization patterns between 60-80%, demonstrating its effectiveness in balancing computational efficiency with healthcare manufacturing quality requirements.