Skip to content

Latest commit

 

History

History
96 lines (69 loc) · 4.07 KB

README.md

File metadata and controls

96 lines (69 loc) · 4.07 KB
Raw

Fleet Adapter Ecobot

This repository contains an RMF fleet adapter developed for Gaussian Ecobot cleaning robots. It has been tested on S40, S50 and S75 variants of robots. The adapter communicates with the robots via REST API calls. For more information on the APIs, please reference this document

Requirements

Setup

  • Clone this repository into a workspace
  • Source Open-RMF
  • colcon build --packages-up-to fleet_adapter_ecobot --cmake-args -DCMAKE_BUILD_TYPE=Release

Run the fleet adapter

Ensure the robot can be pinged.

ros2 run fleet_adapter_ecobot fleet_adapter_ecobot -c CONFIG_FILE -n NAV_GRAPH

To run the fleet adapter with rmf-web, specify the server_uri (-s):

ros2 run fleet_adapter_ecobot fleet_adapter_ecobot -c CONFIG_FILE -n NAV_GRAPH -s ws://localhost:8000/_internal

Test the fleet adapter in "Mock Mode"

The adapter can be tested in mock mode with the help of the TestClientAPI. This class emulates the member functions of EcobotClientAPI.py which calls the rest apis of the robot. This "mock mode" is enabled by providing -tf argument.

Run this example in office world:

ros2 launch rmf_demos office.launch.xml run_fleet_adapters:=0

Then run the ecobot fleet adapter

# Note the addition of "--test_api_config_file" and "-tf"
ros2 run fleet_adapter_ecobot fleet_adapter_ecobot \
    -c src/fleet_adapter_ecobot/configs/robot_config.yaml \
    -n install/rmf_demos_maps/share/rmf_demos_maps/maps/office/nav_graphs/0.yaml \
    -s "ws://localhost:8000/_internal" \
    -tf src/fleet_adapter_ecobot/configs/test_api_config.yaml

Different to the simulation running on gazebo, this TestClientAPI mocks the behavior of the fleet adapter when receives command from RMF. Thus, the script is running on actual system wall time.

Patrol Task

Now try to command robot to move to Pantry

ros2 run rmf_demos_tasks dispatch_patrol -p pantry

Custom Clean Task

Send the robot to clean an area. This custom clean task is created by composing go_to_place and custom perform_action.

ros2 run rmf_demos_tasks dispatch_action -s patrol_D2 -a clean -ad '{ "clean_task_name": "clean_hallway", "clean_mode": "heavy_cleaning" }'

Show overlayed ecobot map

Show overlayed lidar map on rviz2 office map

ros2 launch rmf_demos map_server.launch.py map_name:=ecobot_office tx:=1.33 ty:=0.057 yaw:=-1.598

Docking to Charger

Add a dock_name on a charger waypoint in traffic editor. This will then call the dock() function when the robot is approaching the charger. Note that this is not demonstrated in this demo.

Get the Ecobot to RMF map Transformation with traffic-editor

To get the transformation of the robot map to rmf map, user can add a "floorplan" of a robot map. Then annotate and the corresponding "constraint-pairs", lastly ctrl-T to let traffic-editor calculate the respective transformation.

Specify this transformation to the rmf_transform in the config.yaml file. Note that the value of Y-axis is applied with a -ve.

Then if you wish to configure your custom waypoints in the configs/test_api_config.yaml, you can use rviz to obtain those points in ecobot coordinates. Run this on a separate terminal.

# first run the office map
ros2 launch rmf_demos office.launch.xml run_fleet_adapters:=0
# then run this on a seperate terminal
ros2 run fleet_adapter_ecobot clicked_point_transform -tf 1.33 0.057 -1.598 0.049

Subsequently, select "publish point" on rviz, then click on the respective point on the map. Immediately, the point in rmf and robot coordinate will get printed on clicked_point_transform terminal. These coordinates are helpful during debugging.