Digital Twin (DT) is a rising technology concept for representing the physical-world assets in digitalized form. Real time events can be mapped using DT by monitoring the physical part of an entity and employing both artificial intelligence algorithms and big data analytic tools to create DT models that represent the virtual part. In smart manufacturing, job shop scheduling is affecting the production efficiency. To ensure the maximum utilization of the available resources, a robust job shop scheduling technique is proposed for a poultry feed industrial factory. This has been achieved via two phases. Phase one, a flexible digital twin model is built to simulate the actual physical system of the production lines in the adopted factory. The main objectives of this phase are modeling, monitoring, and graphing the production lines in the factory. In phase two, an efficient job shop scheduling algorithm is presented to schedule and reschedule the jobs among the healthy production lines. The optimization technique used in this phase is a genetic algorithm with enhanced chromosome representation to ensure that all the populations' individuals generated are feasible solutions which reduce the time required to achieve convergence.