A Digital Twin Approach To Job Shop Scheduling: Simulation And Optimization In Anylogic

In complex manufacturing environments such as job shops, machine breakdowns and maintenance activities can significantly disrupt production flow, leading to delays, bottlenecks, and reduced throughput. This thesis presents the development of a digital twin for a job shop system using AnyLogic simulation software to evaluate the effectiveness of fallback routing logic, where jobs are dynamically rerouted to alternative machines when primary machines are unavailable due to scheduled and unscheduled events. The digital twin replicates real-world job shop conditions, incorporating variable job sequences, machine-specific processing times, and both preventive and corrective maintenance schedules. Two scenarios were compared: a baseline configuration with fixed routing with first- in-first-out (FIFO) order and an alternative configuration with fallback logic. Performance was assessed across different KPIs, including makespan, machine utilization, throughput, and average waiting time. Simulation experiments were conducted in a FIFO (First In First OUT) sequence and with downtime events (preventive and corrective maintenance), results were validated through sensitivity analysis and statistical testing. The findings showed that fallback logic reduced makespan by 8.24%, improved throughput by 8.89%, and led to more balanced machine utilization. A two-sample t-test confirmed that the makespan reduction was statistically significant at the 90% confidence level (p = 0.0304). These results suggest that integrating fallback routing into a job shopâ??s digital twin enhances system flexibility and resilience, enabling more adaptive and efficient production control. The study contributes to the growing body of research on digital twin technology and supports its role in enabling smart, disruption-tolerant manufacturing systems aligned with Industry 4.0 objectives.