<p>This paper addresses the challenge of achieving fast and precise positioning control for a robotic manipulator operating under load uncertainty and base vibration. First, detailed dynamic models of nonlinear friction and balance torque are developed in the system. A finite-time disturbance observer (FDO) is then designed to estimate residual uncertainties in real-time. These uncertainties are then compensated for within the control framework, thereby significantly reducing positioning time. Building on this framework, we propose a novel adaptive super-twisting sliding mode controller (ASSMC) that employs an implicit Lyapunov function for online gain adaptation. The controller offers three distinct advantages: automatic gain adjustment via the implicit Lyapunov function, improved robustness to observation errors, and reduced parameter tuning requirements. Comparative experimental results demonstrate that the developed ASSMC exhibits considerable robustness against load uncertainty and base vibration while maintaining high positioning precision. Meanwhile, the compensation for the uncertainties significantly reduces the positioning time from 2.127 s to 0.947 s, greatly enhancing the system’s positioning performance.</p>