In Dual-Resource Constrained Flexible Job Shop Scheduling (DRCFJSP), the common assumption of fixed processing times overlooks the significant impact of worker experience on productivity. This paper introduces a hybrid learning–interference model that replaces simple repetition counts with a more realistic effective experience measure. This measure is novel in several key aspects: it couples position- and duration-based learning, weights experience transfer using both task and machine similarity to capture the human-machine-task fit, models skill decay as a continuous interference process, and incorporates worker heterogeneity. By embedding these dynamics directly into the time formulation, our model provides a theoretically more accurate estimation of processing times. Simulation results indicate that this increased fidelity leads to more effective and robust scheduling solutions for human-centric manufacturing environments.