![]() This way, experiences and results can be interchanged between them. The evolution of the simulator (and of the different prototypes aimed at obtaining one that ressembles the real robot) incorporating characteristics on it, allow us to work “indistinctly” so much in the real robot as in the simulator. Other than that, the errors are also caused by the characteristics of the sensors themselves.This article describes the work accomplished to get a Webots simulator for the Lauron IV robot and how we have solved the difficulties encountered during the project. We need SLAM to give us a more accurate pose. A more intelligent motion pattern will be much better.Īlso, you may notice that the map isn’t very accurate, that is mostly because we only use odometry for the robot pose which is prone to accumulating error over time. ![]() One simple example that I’ve shown here is the autonomous mapping mission for the robot to map its environment to be used for different purposes such as localization.Īs you can see from the videos the random bounce is not efficient because it may take a long time to map even normal-sized rooms. We have seen above that we can use open-source software to learn to develop software for robots and test our software on the simulator.
0 Comments
Leave a Reply. |