Satisfying the increasing water demands while minimizing energy costs is crucial for water supply systems (WSS). Therefore, pump scheduling is critical since the pumps account for most of the operational energy consumption in WSS. The integration of smart predictive digital twins (SPDT) offers a promising approach to this issue, enabling real-time and predictive control of pump operations. However, the Pump Scheduling Problem (PSP), which seeks to balance water demand and operational constraints, must be solved efficiently to support SPDT functionality. The complexity of PSP arises from the nonlinear hydraulic dynamics, fluctuating energy tariffs, and the need to satisfy multiple constraints simultaneously [1]. Traditional methods like Mixed-Integer Linear Programming (MILP) and Nonlinear Programming (NLP) have effectively addressed PSP challenges but struggle with computational inefficiency for large-scale systems. To address these limitations, metaheuristic algorithms and hybrid approaches have shown promise in efficiently exploring solution spaces. A recent study [2] shows that the duty-cycles formulation, combined with local search algorithms like Sequential Least Squares Quadratic Programming (SLSQP), is highly efficient for solving the PSP, offering significant computational and cost-saving benefits. However, it still faces challenges, particularly in handling multiple local optima and scaling large, real-world systems. This communication presents the results of a novel hybrid method, Smart Dynamic Local Search (Smart-DLS), developed to address these challenges. Smart-DLS combines deterministic local search techniques with an intelligent shaking process to overcome the drawbacks of the previously mentioned optimization model. The approach was applied to a real Portuguese WSS for 7 days in a row, and its results were compared to the real WSS operation. The approach outcomes suggest that Smart-DLS enhances the robustness and scalability of PSP solutions, offering an efficient framework to be integrated into the SPDT for optimizing pump operations of real-world WSS.