Production rescheduling with machine breakdown based on data collection from IOT system

Job shop scheduling is critical to manufacturing optimization, encompassing proactive planning and reactive adjustments to unforeseen disruptions. This paper addresses the complex job shop environment characterized by parallel machines within groups, shared machines across processes, and the occasional dedication of machine groups to specific processes due to extended processing times. We propose a novel realtime rescheduling methodology to enhance production efficiency in this dynamic setting. Our approach integrates a Genetic Algorithm (GA) with Adaptive Large Neighborhood Search (ALNS) to effectively explore a vast solution space. We further introduce batch splitting to increase flexibility and responsiveness to disruptions. Rigorous evaluation of diverse problem instances demonstrates the methodology's ability to minimize tardiness and makespan while maintaining computational efficiency. Significantly, our approach effectively repairs schedules disrupted by machine breakdowns, improving delivery times and reducing production costs.