Abstract: With the popularization of industrial robots and the continuous expansion of their application scenarios, how to make them adapt to more application scenarios has become a hot topic in the field of industrial control research. The research focus of the robot picking process is to achieve flexible picking in scenarios such as industrial robot handling and machine tool loading and unloading. The traditional retrieval control method adopts a hybrid force position control strategy that switches between force control mode and position control mode. However, this method requires frequent switching of control modes, which not only increases system complexity but may also reduce system stability and control accuracy. Therefore, this article proposes a force position hybrid control scheme based on hyper spiral control, which eliminates the “strong rigidity” characteristic of industrial robot position control mode and enables industrial robots to work in position control mode to achieve flexible material handling. Firstly, the transformation matrix from the robot’s base coordinate system to the end pose is derived. Secondly, the theoretical torque is calculated by combining the dynamic model and Jacobian matrix. Then the feedback torque is calculated based on the current of each joint and parameters such as the motor and reducer. On this basis, a super spiral controller is designed to improve control robustness while ensuring that the system state and its derivatives converge quickly to the sliding surface. At the same time, a disturbance observer based on second-order low-pass filtering is designed to compensate for the feedforward of the current control loop, enabling the robot to operate stably within a specific pose range. The experiment shows that the force position hybrid control scheme proposed in this paper significantly improves the compliant handling performance of industrial robots without the need for additional torque sensors, effectively reducing the cost and complexity of the system. This method provides a new solution for applications of industrial robots in various flexible picking scenarios, and has important theoretical value and technical significance.