<![CDATA[Production scheduling plays a critical role in manufacturing by determining the sequence and timing of production jobs while balancing multiple operational objectives, including lead time, tardiness, on-time delivery, and machine utilization. Achieving these objectives is increasingly challenging in dynamic manufacturing environments where production efficiency and delivery reliability must be optimized simultaneously. This study presents a descriptive evaluation of intelligent production scheduling using two complementary datasets representing a national steel manufacturing context. The first dataset contains 50 paired production-job records comparing conventional scheduling with optimized scheduling in terms of lead time and tardiness. The second dataset provides an aggregate performance benchmark of five scheduling approaches: First Come First Serve (FCFS), Shortest Processing Time (SPT), Earliest Due Date (EDD), Genetic Algorithm (GA), and Adaptive AI Scheduling. The job-level analysis summarizes paired changes in lead time, reported tardiness, and priority-group performance without interpreting the observations as evidence from a controlled field experiment. Benchmark results indicate that Adaptive AI Scheduling achieved the lowest reported total makes pan of 341.2 hours, compared with 487.3 hours under FCFS, representing a descriptive reduction of 29.98%. The same approach reported a mean tardiness of 1.4 hours, substantially lower than 12.6 hours for FCFS, while achieving 96.0% on-time delivery and 91.7% machine utilization. Across all fifty paired production records, optimized scheduling consistently produced shorter lead times than conventional scheduling. Reported tardiness also declined from three delayed jobs with 16.5 hours of total tardiness under conventional scheduling to one delayed job with 3.9 hours following optimization. Because the datasets do not include a common implementation protocol or experimental controls, the findings describe favorable scheduling patterns rather than causal effects of artificial intelligence. This study contributes a transparent evaluation framework that separates job-level scheduling comparisons from method-level performance benchmarks, providing practical decision support for intelligent manufacturing scheduling assessment and continuous production improvement]]>
Copyrights © 2023
