SLAM (Simultaneous Localization and Mapping)
The goal of this lab project work is to have a deeper knowledge about SLAM (Simultaneous Localization and Mapping). The emphasis on this project is on creating a map of a virtual simulation of a house using SLAM. We’ve used trutleBot3 model name waffel and gazebo house world to simulate a navigation throughout an office compound. The layout of this report lists questions from the main lab document followed by answer and practical discussions that arises during the lab work.
Using the teleoperater node as a controller, the bot3 managed to move and explore unknown areas through the compound. The bot3 model is a burger. I was unable to use a waffel model especially implementing SLAM. One best practice I’ve learned in this task is to set the source with model/setup.bash eventually avoiding to enter model of the robot several times.
A simple navigation Mat-lab-application is used to navigate through the maze. This is done using a probabilistic road map developed in prior labs with few modifications inorder to accommodate obstacle avoidance.The idea behind PRM algorithm is basically, to explore feasible paths around a number of obstacles located in the map. This could be done using mobileRobotPRM object in mat-lab. This app requires a start and a goal position and it then calculates an optimized path towards goal position. The robot movement is controlled using the defined mat-lab application followed by a map builder using Hectors’ method. The map then exported to mat-lab thorough ros-bag. It’s important to note that Occupancy-Map is modified to fit the original map recorded in Rvis. This is due to the color miss-match created when using original occupancy-grid. The result shows the robot controlled using a simple mat-lab application by publishing twist, message in cmdvel topics and reading pose and velocity from Odom topic in real-time. Way-points are defined using the predetermined path created in PRM. Two Matlab-solution is provided for this task in-which the first one is navigation through Operator node and the other is automatic navigation with collusion avoidance. The latter is presented above as figure two:-
https://www.youtube.com/watch?v=_V5AzZ8aMZ4&t=77s https://www.youtube.com/watch?v=_V5AzZ8aMZ4&t=77s
Solomon G
Special thanks to resources on https://emanual.robotis.com/docs/en/platform/turtlebot3/overview/,
- Univestiy West - Part of school project