A simulation meta repository for running the SMADS system using the UT_multirobot_sim from UT_AMRL.
This is a one-stop repository to simulate a robot. Once this simulation is running, the SMADS system can be deployed just as it would on a real robot.
- Clone locally in the directory of your choice:
$ git clone https://github.com/UTSMADS/ut_smads_simulation.git --recursive
-
Add these packages to your ROS_PACKAGE_PATH. You should only need to do this once
$ ./SetupPath
-
Install dependent packages
$ ./InstallPackages
-
Manually install ceres according to its install guide
In the root of the ut_smads_simulation
repo folder, simply run
$ make
In the root of the ut_smads_simulation
repo folrder, simply run
$ ./run_smads_sim.bash
this will spawn a number of terminal windows, a roscore and an RViz visualization. If you don't want to, or can't, visualize locally, you can edit the above script and turn off the visualization flag. Similarly, if a roscore is already running, that flag can also be adjusted in the script.
The relevant files to change maps, navigation or localization parameters live in the config
directory.
These files can be modified or cloned for other robot configurations.
To change the map, you will need to modify the following files:
config/enml/ut_jackal.lua
under the fieldmap_name
config/sim/default_init_config.lua
under the fieldmap_name
Note that the starting_loc_*
listed under config/enml/ut_jackal.lua
should match that of the start_poses
under config/sim/default_init_config.lua
for the simulation to work 'out of the box'. Otherwise, you will need to give ENML a new pose estimate to match the robot's actual position in the simulator.