This paper presents an online disturbance-compensated adaptive sliding mode control strategy for a six-degree-of-freedom industrial robotic manipulator operating under dynamic uncertainties and external disturbances. The proposed method employs a wavelet neural network to approximate uncertain dynamics and an adaptive estimator to handle unmeasurable external disturbances, enabling real-time compensation for both uncertainty sources. By integrating these components into a unified sliding mode control framework, the controller adapts dynamically to internal variations and external influences without requiring prior knowledge of their bounds. This approach mitigates chattering in the control signal, resulting in smoother actuator responses and improved tracking performance. Simulation results demonstrate that the strategy achieves accurate trajectory tracking, strong robustness to disturbances, and significantly reduced control signal oscillations. These results highlight the effectiveness of combining wavelet-based approximation and real-time disturbance estimation to enhance robustness and smoothness of sliding mode control in industrial robotic systems.