This a log file for the introductory course of ROS, held at ETH, RSL department. In the following chapters I will cover the main points of each exercise given at the end of each lecture.
The aim of this repo is to demonstrate what I have learned during this course, giving also the possibility to share my solutions with other students.
. Naming convention for the package: it should start with a lower case.
. A reference & is added when we want to just reference to the variable address, without creating a new variable (occuping more memory), when I have a reference I can modify directly the source variable.
. A common usage of const is to guard against accidentally changing a variable, especially when it is passed-by-reference as a function argument. For this reason we have: void function(const std::vector<int> &v)
. A package can have multiple nodes inside, when creating a new node (main.. init ..ecc.) pay attention to add an executable to CMakeList.txt following this template add_executable("executable name" "executable files source") and target_link_libraries("executable name" ${catkin_LIBRARIES}).
. When subscribing to a new topic, we just need to choose a new topic name, the topic will be generated automatically.
1.
Use command catkin_create_pkg to create a new package (add dependencies). We then open CMake ans package.xml files and uncomment the lines needed (lecture 2, slide 5-7). Next step is to create .cpp/.h files. First we build the project, then we import to Eclipse, now we are ready to create the new files.
include/ROS_ETH_course/ControllerHeader.hpp is the header file, it contains all the declarations, useful when dealing with a multi file program.
src/ROS_ETH_course_node.cpp contains the main loop (node file).
src/ControllerMain.cpp contains the main program with all the instruction. This program is called by ROS_ETH_course_node.cpp.
4.
When creating the subscriber, pay attention to pointer/reference in the line, use this pointer when pointing to a function
5.
Create a .yaml file with the specified parameters, in the launch file specify to load these parameters in the node. Then in the actual program, we get the desired params with NodeHandle.getParam.
6.
When addressing to a member of a class that is a pointer, we use the -> notation, instead when referring directly to a class object we use the . notation. Moreover if we want to refer directly to a class member (not the object) we have use :: notation.
Init a vector using std::vector and then go through (for loop) all the ranges to find the minimum one.
7-8. When passing a node in the .launch file, pay attention to the type, this is the name of the file within the package we want to launch.
9. Before calling a package remember to build it.
After setting the first time the configuration in RViz, we can set it as default config by typing ctrl+s. This way next time we start RViz everything will be environment will be ready.
1.
Insert an arg parameter to to link directly to the name of the desired world file.
3-4. Create a new publisher in the main program, with the type of message we want to send and the topic where we want to publish. After that in the callback function we have compute the angular position of the pillar with respect to husky. We combine Min Distance and Pillar Angle to create a P controller that increase velocity and steers. Finally we add the computed values to the Twist message, before publishing it.
7. When computing a transformation with tf, put it inside a try statement, otherwise at the startup, when not all the tf are loaded, a fatal error will occur.
When trasforming using transformPoint it is important to assign the frame_id to the point that needs to be converted.
In the terminal type rosrun tf view_frames for generating a .pdf file with the mapping of the all the coordinate systems.
2.
Before installing rqt_multiplot we need to install the required pre-requisites.
4.
After checking the subscription/publication of the bag file, we need to add to the launch file all the nodes required to have the same configuration, check in the reference project husky_control/launch file `control.launch`` which nodes have been launched.
5.
Normally, the ROS client libraries will use your computer's system clock as a time source, also known as the "wall-clock" or "wall-time" (like the clock on the wall of your lab). When you are running a simulation or playing back logged data, however, it is often desirable to instead have the system use a simulated clock so that you can have accelerated, slowed, or stepped control over your system's perceived time. For example, if you are playing back sensor data into your system, you may wish to have your time correspond to the timestamps of the sensor data. In the .launch file set <param name="use_sim_time" value="true"/>.
When calling the bag file, remember to set the option --clock to send the bag-s clock data to the /clock topic.
6.
When calling the node robot_state_publisher first we need to call the robot_description from the parameter server. This can be found in the husky_description package (optenv substitute the value of a environment variable if set).
1.
We create a .cpp file with a server in it, this server has a type msg of std_srvs, and communicate using bools (request, response).
In the callback function of the service, we send a response for received request and we publish in a topic passing a variable (bool) for starting/stopping the motion.
Once the server node is running, we use the command rosservice call "service_name" "request value" for passing the request variable.
B. Analysing the imu data, I noticed a significan bump in the acceleration in the z coordinate when hitting the pillar.
For implementing this point we need to create a service client. We subscribe to the imu/data topic, type sensor_msgs/Imu.h. Once the acceleration is higher than a fixed threshold, we send a call for the service implemented in point 1.
Before launching remember to add server and client to the launch file (two distinct nodes), and to the Cmake file as executable.