Robot Assisted Manufacturing

The robot assisted manufacturing project currently focused on implant fabrication built in ROS2 foxy.

Packages

Current packages:

ram_gripper_control - simulated gripper controller
ram_interfaces - custom interfaces for ram project
ram_motion_planning - components to plan trajectories for toolpaths
ram_moveit_config - configuration files for moveit
ram_support - models and URDF for the core components of the project (robots, tables, etc)
ram_tooling_support - assistance nodes for handling (currently) toolpaths and stock material

Running the system

This launches RVIZ, move_group node, sim iiwa controller, sim gripper controller, toolpath handler and stock handler. Now includes some arguments that can be set (just for the ros control side at the moment)

Defaults to real robot now

ros2 launch ram_motion_planning ram_control.launch.py

Add robot_port:=30200 robot_ip:=192.170.10.2 to set the robot ip and port (make sure you also set that on the robot controller)

Use Sim robot

ros2 launch ram_motion_planning ram_control.launch.py real_manipulator:=false

This launches the moveit movegroup interface for fully constrained toolpaths (6DoF) quite simple at the moment or the ompl constrained (also very simple)

ros2 launch ram_motion_planning toolpath_planner.launch.py  

ros2 launch ram_motion_planning ompl_toolpath_planner.launch.py

When both launch files are running (wait for rviz to load), the rqt gui or these service calls can be used to:

load the toolpath from the toolpath handler
Compute the part to tool transform
Plan to the approach pose
Plan the toolpath with the retreat pose added

toolpath execute will send the toolpath trajectory to the robot controller

ros2 service call /toolpath_planner/toolpath_setup std_srvs/srv/Trigger {}  
ros2 service call /toolpath_planner/toolpath_execute std_srvs/srv/Trigger {}

Docker

Helper scripts

You can use the full commands below but also use the helper scripts in the .docker folder.
Currently implemented helpers (As always look through the scripts first before running them)
All should be run from the root directory of the repo.
.docker/build_image.sh -- Build the image tagged as gdwyer/ram:-amd64
.docker/run_user.sh -- Runs the container as the user with x server etc (same as the command in the Run section below)
.docker/run_user_nvidia.sh -- Same as above but with access to a nvidia GPU added (Needs nvidia docker2 to be installed)
.docker/run_control.sh -- Runs the container as root with the kuka driver added to the repo but not built, container will delete itself after exit.

Build

docker build --pull --rm -f ./.docker/Dockerfile  -t gdwyer/ram:<branch>-<platform> .

Run

docker run -it \
    --user=$(id -u $USER):$(id -g $USER) \
    --group-add sudo \
    --env="DISPLAY" \
    --env=QT_X11_NO_MITSHM=1 \
    --workdir="/dev_ws" \
    --volume="/home/$USER:/home/$USER" \
    --volume="/etc/group:/etc/group:ro" \
    --volume="/etc/passwd:/etc/passwd:ro" \
    --volume="/etc/shadow:/etc/shadow:ro" \
    --volume="/etc/sudoers.d:/etc/sudoers.d:ro" \
    --volume="/tmp/.X11-unix:/tmp/.X11-unix:rw" \
    --net=host \
    gdwyer/ram:<branch>

If on a computer with nvidia-docker2 installed add --gpus 'all,"capabilities=compute,display,graphics,utility"' before the image. Full docker info here

Hardware Support

Iiwa

WARNING this currently only works if you run as a user and have ssh keys setup for the Kuka_iiwa_ROS2_drive repo /dev_ws/src/ram/.docker/add_kuka_driver.sh

grdwyer / Robot-Assisted-Manufacturing

readme