Mobile robot system, mobile robot, and method of controlling the mobile robot system

A mobile robot comprising:

• a main body;
• a driving portion including at least one wheel configured to move the main body;
• a sensor portion configured to recognize objects existing in a moving space;
• a controller configured to transmit information about moving space to an external server and receive, from the server, first traveling information about the moving space created based on the information about the moving space;
• and
• the controller is configured to create second traveling information including a traveling path in the moving space based on the first traveling information and information about the recognized objects in the moving space, and to control the driving portion to move the main body according to the traveling path in the moving space included in the created second traveling information,
• wherein the first traveling information is created based on information about the moving space received from at least one another external robot different from the mobile robot;
• and
• wherein the first traveling information is created using deep learning based on the information about the moving space using a Deep Neural Network (DNN), stored on the server, via repeated convolutions and sub-samplings, and the controller is configured to perform deep learning based on the first traveling information and information about the recognized objects in the moving space to create the second traveling information using a DNN stored on the mobile robot,
• wherein the information about the moving space includes information created based on a simulation of the mobile robot or the external robot traveling on the moving space, and
• wherein a learning level of the Deep Neural Network (DNN) stored on the server is higher than that of the DNN stored on the mobile robot.

A method of controlling a mobile robot comprising a main body, a driving portion including at least one wheel configured to move the main body, a sensor and a controller, comprising:

• receiving, by a server, information about a moving space from at least one external robot;
• creating, by the server, first traveling information about the moving space based on the information about the moving space from the external robot;
• recognizing, using the sensor, objects in the moving space;
• creating, by the controller, second traveling information including a traveling path in the moving space based on the first traveling information and information about the recognized objects in the moving space;
• and
• controlling, by the controller, the driving portion to move the main body of the mobile robot according to the traveling path in the moving space included in the second traveling information, the mobile robot being different than the external robot,
• wherein deep learning is performed based on the information about the moving space to create the first traveling information using a Deep Neural Network (DNN) stored on the server, via repeated convolutions and sub-samplings, and
• wherein deep learning is performed based on the first traveling information and information about the recognized objects in the moving space to create the second traveling information using a DNN stored on the mobile robot,
• wherein the information about the moving space includes information created based on a simulation of the mobile robot or the external robot traveling on the moving space, and
• wherein a learning level of the Deep Neural Network (DNN) stored on the server is higher than that of the DNN stored on the mobile robot.