Method of performing cloud slam in real time, and robot and cloud server for implementing the same

A cloud server for performing cloud simultaneous localization and mapping (SLAM) in real time, the cloud server comprising:

• a storage device configured to store a map;
• a communication device configured to communicate with a plurality of robots;
• and
• a controller configured to generate a map based on sensor data transmitted by the plurality of robots and configured to generate a global pose or a local map patch for each of the plurality of robots.

A method of performing cloud simultaneous localization and mapping (SLAM) in real time, the method comprising:

• acquiring sensor data from a first sensor of a robot;
• extracting, by a controller of the robot, a feature from the sensor data and performing local mapping to generate a local pose;
• transmitting, by a communication device of the robot, the sensor data and the extracted feature to a cloud server;
• generating, by a controller of the cloud server, a local map patch based on the sensor data and the extracted feature transmitted from the robot;
• transmitting, by a communication device of the cloud server, the local map patch to the robot;
• generating, by the controller of the robot, loop constraints based the received local map patch and subsequently generating a first global pose;
• and
• controlling, by the controller of the robot, movement of the robot based on the generated first global pose.