Method for learning robot task and robot system using the same

A robot comprising:

• a first controller configured to receive a teaching command of a task, from a haptic controller, the haptic controller configured for a user to manipulate the robot to teach the task, the haptic controller configured to output haptic feedback based on force information from the robot such that the user feels the force information at the robot at a time of the robot performing the task, while teaching of the task to the robot is performed;
• a driver having a multi-joint, the driver configured to perform a driving according to the teaching command under control of the first controller;
• a sensor configured to sense first information related to a task environment of the robot and second information related to a driving state of the robot and store, in a database, the first information and the second information as time-series data with the teaching command, while the driver is driven under the control of the first controller;
• and
• a second controller configured to drive the driver by controlling the first controller,
• wherein the robot is configured to perform learning of the task through a machine learning by using the first information and the second information stored in the database,
• wherein the second controller is configured to generate a control command with respect to the first controller such that the robot autonomously performs the task based on the learning of the task, and
• wherein the first information includes at least one of a capturing image, a distance from an object or a temperature of the task environment.

A method for learning a robot task, comprising:

• performing teaching of a robot task based on an external manipulation of a haptic controller, the haptic controller configured for a user to manipulate a robot to teach the robot task;
• transmitting a driving command corresponding to the external manipulation to the robot by the haptic controller, and outputting haptic feedback by the haptic controller based on force information from the robot such that the user feels the force information at the robot at a time of the robot performing the robot task, while teaching of the robot task to the robot is performed;
• sensing first information related to a task environment of the robot and second information related to a driving state of the robot by a sensor, while the robot performs the robot task according to the driving command;
• storing, in a database, the first information and the second information sensed by the sensor as time-series data together with a teaching command;
• and
• performing learning of the robot task through a machine learning by using the first information and the second information stored in the database,
• wherein the first information includes at least one of a capturing image, a distance from an object or a temperature of the task environment.