This paper presents a new nonlinear disturbance observer (NDO) for robotic manipulators, designed to overcome the limitations of existing disturbance observers (DOs) that are typically based on linear system techniques. The NDO is derived to address various applications such as friction compensation, independent joint control, sensorless torque control, and fault diagnosis. The global exponential stability of the NDO is guaranteed by selecting design parameters that depend on the maximum velocity and physical parameters of the robotic manipulator. The performance of the NDO is demonstrated through simulations and experiments on a two-link robotic manipulator, where it successfully estimates and compensates for friction. The results show that the NDO improves the tracking performance and reduces steady-state errors, making it a valuable tool for enhancing the robustness and reliability of robotic systems.This paper presents a new nonlinear disturbance observer (NDO) for robotic manipulators, designed to overcome the limitations of existing disturbance observers (DOs) that are typically based on linear system techniques. The NDO is derived to address various applications such as friction compensation, independent joint control, sensorless torque control, and fault diagnosis. The global exponential stability of the NDO is guaranteed by selecting design parameters that depend on the maximum velocity and physical parameters of the robotic manipulator. The performance of the NDO is demonstrated through simulations and experiments on a two-link robotic manipulator, where it successfully estimates and compensates for friction. The results show that the NDO improves the tracking performance and reduces steady-state errors, making it a valuable tool for enhancing the robustness and reliability of robotic systems.