Normally, Industrial robots must be operated where the environment is contrivedto suit the robot. The robots have less use in the applications where the workingenvironment and the object position can not be accurately specified. This limitationis due to the lack of sensory capability. Force and vision sensors can be used toimprove the flexibility of the robot since they mimic the human sense of force andvision respectively. The robots are able to interact with its environment using thesecapabilities. In this research, the vision system is used to navigate the manipulator hand toa target object. The camera is in an eye-in-hand configuration and is able to capturethe target position relative to the manipulator hand. Force sensor is alsos equippedto the robot hand. The task is to control force and visual parameters simultaneously,i.e., the target is in the middle of the image and also has the specified touchingforce, a 3 degrees of freedom task. The position of the target in the worldcoordinate is unknown. A new technique called Learning Control is developed to complete the task. Thistechnique determines the Jacobian from the robot movement and uses it to determinethe direction for updating joint angles. Since this technique does not require systemmodeling, it is applicable to any robots. Furthermore, this technique works on jointcommand and thus is able to implement on any industrial robot controller. Weperformed several experiments to validate the proposed method. The computersimulation is used to analyze the performance in various conditions.