In this paper, we propose novel driving methods for autonomous driving of robots in order to overcome the disadvantages of existing guidance systems applied to indoor autonomous mobile robots.
In the grid pattern tracking autonomous driving method, the robot recognizes the location and determines the driving direction by using grid pattern information such as tiles installed on the floor surface of the indoor space only with information from the camera attached to the robot without additional construction or addition of devices.
The robot that tracks grid pattern can determine the current location, using a character matrix map, plan a route to the destination, and then perform autonomous driving. Furthermore, it is possible to increase the accuracy and freedom of indoor autonomous driving robots by utilizing cameras attached to robots and optical character recognition techniques to recognize text information of guide signs installed in indoor driving spaces.
We developed a differential-driven type mobile robot to evaluate the performance of indoor autonomous driving robots that combine grid pattern tracking and character matrix map proposed in this paper. The robot built for experiments used only the vision sensor as an information gathering device for autonomous driving.
Experiments conducted to evaluate the performance of the indoor autonomous mobile robot that mixes grid pattern tracking and character matrix map confirm that the autonomous driving method by grid pattern tracking is highly accurate with only image information obtained from cameras attached to the robot. In addition, it can be confirmed that the indoor autonomous driving robot, which combines grid pattern tracking and character matrix map, has superior degrees of freedom, accuracy, expandability and flexibility compared to the existing guidance system.
The indoor autonomous driving method of a robot that combines grid pattern tracking and character matrix map proposed by us is simpler and more economical than the existing method because it enables location recognition and driving control of a mobile robot only with information from the camera attached to the robot. Moreover, the indoor autonomous driving method of a robot that mixes grid pattern tracking and character matrix map can overcome the disadvantage of wireless guidance system, which is the complexity of driving route generation, and can improve the expandability and flexibility of the wire guidance system. Therefore, It can be said that it is an autonomous driving method for indoor mobile robots, which has relatively many advantages compared to the existing guidance system.