Skip to content

quick_start_guide.md

Latest commit

 

History

History
148 lines (116 loc) · 3.52 KB

quick_start_guide.md

File metadata and controls

148 lines (116 loc) · 3.52 KB
Raw

Quick Start Guide

This chapter contains a quick start guide for the AIP LLM application.
Please use the GitHub commit status with the tag v1.0.

Disclaimer:
As prerequisites, it requires a running plc for safety , vaccuum system and a running EKI connection on the KUKA SmartPad. If you require more information, please check out the rest of the documentation.

Terminal 1 - Bring Up

  1. Start the docker container

    cd ~/aip_ros2_ws/src/aip_bringup
    . start_docker.sh

  2. Launch the corresponding nodes for aip_bringup

    ros2 launch aip_bringup aip.launch.py use_fake_hardware:=false

Note:

  • The EKI Hardware interface must be running on the KUKA SmartPad
  • If the safety system detects any potential risk and stops the system, the system needs to be acknowledged and restarted (EKI programm on KUKA SmartPad & Terminal for aip_bringup)
  • If use_fake_hardware:=true then the bosch gripper controller node won't be started.
  • This can be used if running AIP virtually only in Rviz
  1. (Optional) Reset the gripper positions 
    # connect to container
docker exec -it aip_bringup bash 

# reset gripper positions
ros2 service call /open_gripper aip_interfaces/srv/MoveGripper "cylinder_ids: {cylinder_ids: [1,2,3,4]}"

Terminal 3 - Roboception driver

  1. Start the docker container 
    cd ~/aip_ros2_ws/src/roboception_driver
    . start_docker.sh

The driver will be started automatically.

Terminal 4 - Object Detector Tensorflow (ODTF)

  1. Start the docker container

    cd ~/aip_ros2_ws/src/object_detector_tensorflow
    . start_docker.sh --no-gpus

    If you want to use the the GPU hardware performance, please delete the flag --no-gpus from the shell command.

  2. Launch the detection nodes

    ros2 launch object_detector_tensorflow detection.launch.py

To show the odtf result manually, it is possible to use the client with:

	ros2 run object_detector_tensorflow client

Terminal 5 - Grasp Planning

  1. Start the docker container

    cd ~/aip_ros2_ws/src/aip_grasp_planning
    . start_docker.sh

  2. Launch the grasp_planning nodes

    ros2 launch aip_grasp_planning grasp_planning.launch.py

    If running the launch file via VS-Code debugging functionality, please additionally use:

    ros2 launch point_transformation point_transformation.launch.py

Terminal 6 - LLM Back End and launch Website

  1. Start the docker container with autostart LLM

    cd ~/aip_ros2_ws/src/LLM_Scene_Docker
    . start_docker.sh

  2. Launch the website node

    ros2 launch pkg_website_llm launch_all_services.launch.py

How to access the website for user interaction?

Terminal 7 - Packing Planning

  1. Start the docker container

    cd ~/aip_ros2_ws/src/aip_packing_algorithm 
    . start_docker.sh

  2. Launch the pack server node

    ros2 run pkg_pack_node pack_server

Terminal 8 - Coordinator

  1. Start the docker container

    cd ~/aip_ros2_ws/src/aip_coordinator 
    . start_docker.sh

  2. Start the Behavior Tree sequence via the launch file

    ros2 launch aip_coordinator aip.launch.py

RESET Application

restart of some specific nodes is necessary:

  • aip_coordinator
  • aip_packing_algorithm
  • (in simulation) aip_bosch_gripper