GithubHelp home page GithubHelp logo

cabot's Introduction

License: MIT

CaBot

CaBot (Carry on Robot) is an AI suitcase to help people with visual impairments travel independently. Can you imagine walking around at airports without vision? It's a huge open space, and there are lots of things and people so that it is hazardous for them to walk around at airports. see project detail

CaBot v2

CaBot v2 has been migrated to ROS2 (humble). A refactoring has changed its structure (Dec. 2023), previous development branch is dev-ros2. After the refactoring, most of ROS2 packages for the CaBot are managed under sub repositories. Please check those repositories for the details.

Sub repositories

Hardware (example)

  • Robot frame + handle
  • LiDAR
    • Velodyne VLP-16
  • Stereo Camera(s)
    • 1 RealSense camera (D435)
    • 2~ RealSense cameras (with a PC or NUC + Jetson cluster)
  • Motor Controller
    • ODrive Motor Controller v3.6 (Firmware v0.5.1)
  • Micro Controller (Handle and sensors)
  • GNSS Receiver (optional)
    • GNSS board with ublox ZED-F9P module
  • Processor
    • PC with NVIDIA GPU (ZOTAC Magnus EN72070V)
    • NUC (Ruby R8) + Jetson Mate (multiple Jetson Xavier NX)

Localization

Tested Environment

  • PC
    • Host Ubuntu 20.04
    • Docker v20
    • docker-compose v2.20.2
  • Jetson
    • Host Ubuntu 20.04 (Jetpack 5.1)
    • See jetson for detail

Setup

  • import third-party repos by using vcstool
    pip3 install vcstool # if you don't have vcs
    ./setup-dependency.sh
    
  • run all scripts in tools based on your requirements
    cd tools
    ./install-service.sh               # need to install to configure system settings
    ./install-docker.sh                # if you need docker
    ./install-arm-emulator.sh          # if you build docker image for Jetson
    ./install-host-ros2.sh             # if you watch system performance or debug
    ./install-realsense-udev-rules.sh  # if you use realsense camera
    ./setup-display.sh                 # for display connections from docker containers
    ./setup-usb.sh                     # if you run physical robot
    

Prepare Docker Images

Pulling from dockerhub

  • pulling docker containers
    ./manage-docker-image.sh -a pull -i "ros2 localization people people-nuc ble_scan" -o cmucal -t ros2-dev-latest
    
  • build docker workspace and host workspace
    ./build-docker.sh -w -o
    

Build Docker Images from scratch

  • build docker containers (at top directory)
    ./build-docker.sh -p -i -w -o
    

Launch

  • Run containers. Please configure the .env file before launching
    ./launch.sh                 # for robot
    ./launch.sh -s              # for simulator
    ./launch.sh -c nuc          # for robot with nuc machine (need to configure CABOT_JETSON_CONFIG)
    ./launch.sh -c rs3          # for robot with 3 realsense configuration
    ./launch.sh -c rs3-framos   # for robot with 3 FRAMOS configuration
    
    other options
      -s          simulation mode
      -d          do not record
      -r          record camera
      -p <name>   docker compose's project name
      -n <name>   set log name prefix
      -v          verbose option
      -c <name>   config name (default=) docker-compose(-<name>)(-production).yaml will use
      	if there is no nvidia-smi and config name is not set, automatically set to 'nuc'
      -3          equivalent to -c rs3
      -M          log dmesg output
      -S          record screen cast
      -t          run test
    
  • (optional) Run the gnss container before running launch.sh if you use a gnss receiver
    docker-compose -f docker-compose-gnss.yaml up
    

.env file

  • Required settings

    CABOT_MODEL          # robot model (default=) to determine which launch/urdf to use
    CABOT_SITE           # package name for cabot site (default=)
    CABOT_TOUCH_PARAMS   # touch sensor parameter for cabot-arduino handle (default=[128,48,24])
    RMW_IMPLEMENTATION=rmw_cyclonedds_cpp   # need to use cyclone dds due to performance issue
    
  • Required settings for 3 Realsense configuration

    • _X should be replaced with _1, _2, or _3 for each realsense
    CABOT_REALSENSE_SERIAL_X      # serial number of realsense
    CABOT_CAMERA_NAME_X           # camera name and camera should be at '<name>_link' (TF)
    
  • Required settings for 2 odrive configuration (cabot3 model)

    CABOT_ODRIVER_SERIAL_0  # serial number of odriver (left wheel)
    CABOT_ODRIVER_SERIAL_1  # serial number of odriver (rigth wheel)
    
  • Settings for the configuration using Jetson (experimental)

    • This will up people docker container on each specified jetson (by IP address or hostname).
    • Each jetson should connect to a Realsense
    • Each jetson should be ssh identification login enabled (without password) from the main machine
    • Each jetson's .env file should be configured proper CYCLONEDDS_URI setting
    CABOT_JETSON_USER    # User name to login jetson (default=cabot)
    CABOT_JETSON_CONFIG  # Space separated config for muliple jeston/realsense
      #
      # "<<Mode>:<HOST>:<Name>[ <Mode>:<IP Address>:<Name>]*"
      # Mode: D - detection
      #     : T - tracking (only for testing, the PC will launch tracking processes)
      # HOST: IP address or host name of the jeston
      # Name: Used for camera name space "<Name>" and camera TF link "<Name>_link"
      #
      # ex) "D:192.168.1.50:rs1 D:192.168.1.51:rs2"
    CYCLONEDDS_URI=/home/developer/cyclonedds_spdp.xml   # need to use cyclone dds SPDP message option for multiple hosts
    
    • Run following command on each jetson to enable service for changing DDS settings
    tools/install-jetson-service.sh
    
  • Troubleshooting settings when cabot simulation does not work properly as shown below

    • List of problems
      • Map does not appear
      • Map disappears
      • Costmap does not appear
      • Green pathes do not appear (, but blue nodes appear)
      • Robot model does not appear and white model is shown instead
      • Robot model blinks
      • Button "Navigate to here" does not work
      • Button "Navigate to here" works but global path is being loaded forever. Navigation will not start.
      • Requires multiple presses on "Navigate to here" to start navigation
      • Positional drift right after reaching the destination
    • Solution
      • Pre-Requisites before implementing the following workaround
        1. Execute tools/install-service.sh
        2. Verify that RMW_IMPLEMENTATION is set correctly in .env
        3. Verify that ROS_DOMAIN_ID is set to a unique value in .env, if there are other PCs running ROS 2 on the same network
      • The problem is related to networks and DDS settings.
      • Requires the settings below.
        1. Edit .env
        CYCLONEDDS_URI=/home/developer/cyclonedds_spdp.xml
        
        1. Create xml file cabot-navigation/docker/home/cyclonedds_spdp.xml. Replace INSERT.YOUR.IP.ADDRESS with your active IP address.
        <?xml version="1.0" encoding="UTF-8" ?>
        <CycloneDDS xmlns="https://cdds.io/config" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"    xsi:schemaLocation="https://cdds.io/config https://raw.githubusercontent.com/eclipse-cyclonedds/cyclonedds/master/etc/cyclonedds.xsd">
          <Domain>
            <General>
              <Interfaces>
        	<NetworkInterface address="INSERT.YOUR.IP.ADDRESS" />
              </Interfaces>
              <AllowMulticast>spdp</AllowMulticast>
            </General>
            <Discovery>
              <Peers>
        	<Peer address="INSERT.YOUR.IP.ADDRESS"/>
              </Peers>
            </Discovery>
          </Domain>
        </CycloneDDS>
        
  • Optional settings for ./launch.sh options in service

    CABOT_LAUNCH_CONFIG_NAME    # "", "nuc", "rs3"
    CABOT_LAUNCH_DO_NOT_RECORD  # 1/0
    CABOT_LAUNCH_RECORD_CAMERA  # 1/0
    CABOT_LAUNCH_LOG_PREFIX     # string, default=cabot
    
  • Optional settings

    CABOT_BLE_VERSION    # BLE version setting (default=1)
                         # 0: do not use BLE, 1: use BLE server on ROS, 2: use BLE server outside of ROS (needs to setup)
    CABOT_LANG           # cabot language (default=en)
    CABOT_OFFSET         # offset size (default=0.25)
    CABOT_FOOTPRINT_RADIUS # robot radius size (default=0.45)
    CABOT_INIT_SPEED     # specify maximum robot speed at startup, leave empty to restore the last speed
    CABOT_MAX_SPEED      # you can change max_speed only when CABOT_MODEL is cabot2-gtm (defalt=1.0)
    CABOT_USE_GNSS       # to use GNSS fix for localization (default=0)
    CABOT_ANNOUNCE_NO_TOUCH # announce when the reason robot is stopped is NO_TOUCH(default=false)
    CABOT_SIDE           # left: user stands on the right, right: user stands on the left
    CABOT_IMU_ACCEL_BIAS # set parameters to adjust IMU linear acceleration (default=[0.0,0.0,0.0])
    CABOT_IMU_GYRO_BIAS  # set parameters to adjust IMU angular velocity (default=[0.0,0.0,0.0])
    CABOT_VIBRATOR_TYPE  # you can change vibrator type of handle only when CABOT_USE_DIRECTIONAL_INDICATOR is true (default=1)
                         # 1: ERM (Eccentric Rotating Mass), 2: LRA (Linear Resonant Actuator)
    CABOT_USE_DIRECTIONAL_INDICATOR   # to use directional indicator of handle (default=false)
    CYCLONEDDS_NETWORK_INTERFACE_NAME # to specify network interface name for Cyclone DDS
    ROS_DOMAIN_ID        # to specify ROS domain ID; set this value when you use multiple ROS2 systems on the same network
    __NV_PRIME_RENDER_OFFLOAD  # to use NVIDIA GPU for rendering; set to 1 if needed
    __GLX_VENDOR_LIBRARY_NAME  # to use NVIDIA GPU for rendering; set to "nvidia" if needed
    
  • Options for GNSS

    # gnss node
    GNSS_NODE_START_AT_LAUNCH     # whether to launch gnss node at cabot software launch (default=1)
    # ntrip client node
    NTRIP_CLIENT_START_AT_LAUNCH  # whether to launch ntrip client node at cabot software launch (default=0)
    NTRIP_CLIENT                  # select ntrip client to launch from rtlkib, str2str_node, or ntrip_client (default=ntrip_client)
    # options common to str2str_node and ntrip_client
    NTRIP_HOST                    # hostname or IP address of the ntrip server to connect
    NTRIP_PORT                    # port to connect to the server (default=2101)
    NTRIP_MOUNTPOINT              # mountpoint to connect to the server
    NTRIP_AUTHENTIFICATE          # whether to authentificate with the server
    NTRIP_USERNAME                # username to authentificate with the server
    NTRIP_PASSWORD                # password to authentificate with the server
    # options for str2str_node
    NTRIP_STR2STR_RELAY_BACK      # relay back output stream to input stream in str2str_node
    
  • Options for debug/test

    CABOT_GAMEPAD              # (default=gamepad) gamepad type for remote controll (ex. PS4 controller)
                                                 pro (Nintendo Switch Pro controller)
    CABOT_USE_HANDLE_SIMULATOR # to use handle simulator (default=0)
    CABOT_REMOTE_USE_KEYBOARD  # to use teleop twist keayboard in the remote control mode (default=false)
    CABOT_REMOTE_USE_IMU       # to use imu in the remote control mode (default=false)
    CABOT_SHOW_GAZEBO_CLIENT   # show gazebo client (default=0)
    CABOT_SHOW_ROS1_RVIZ       # show ROS1 rviz (default=0)
    CABOT_SHOW_ROS2_RVIZ       # show ROS2 rviz (default=1)
    CABOT_SHOW_ROS2_LOCAL_RVIZ # show ROS2 local navigation rviz (default=0)
    CABOT_SHOW_LOC_RVIZ        # show ROS1 localization rviz (default=1)
    CABOT_SHOW_PEOPLE_RVIZ     # show ROS1 people rviz (default=0)
    CABOT_SHOW_ROBOT_MONITOR   # show robot monitor (default=1)
    CABOT_RECORD_ROSBAG2       # record BT log, controller critics evalation into rosbag2 (default=1)
    CABOT_CAMERA_RGB_FPS       # camera RGB fps (default=30)
    CABOT_CAMERA_DEPTH_FPS     # camera depth fps (default=15)
    CABOT_CAMERA_RESOLUTION    # camera horizontal resolution (default=1280)
                               # need to use 848 or 640 if you use 3 realsense on a PC
    CABOT_DETECT_PEOPLE_FPS    # diagnostic PeopleDetect and CameraInput fps (default=15.0)
    CABOT_PEOPLE_TRACK_FPS     # diagnostic PeopleTrack fps (default=30.0)
    CABOT_DETECT_VERSION       # 1-9 (default=3)
                               # 1: python-opencv, 2: cpp-opencv-node, 3: cpp-opencv-nodelet
                               # 4: python-mmdet, 5: cpp-mmdet-node, 6: cpp-mmdet-nodelet
                               # 7: python-mmdet-seg, 8: cpp-mmdet-seg-node, 9: cpp-mmdet-seg-nodelet
    CABOT_DETECT_PEOPLE_CONF_THRES  # confidence value threshold to detect people (default=0.6)
    CABOT_DETECT_PEOPLE_CLEAR_TIME  # time to clear tracked people from map (default=0.2)
    CABOT_PUBLISH_DETECT_IMAGE # publish people detection image only for debug purpose (default=0)
    CABOT_USE_ROBOT_TTS        # use TTS service '/speak_robot' to let PC speaker speak (default=0)
                               # this function is not used now, but maybe used in some scenario
    TEXT_TO_SPEECH_APIKEY      # IBM Cloud Text to Speech Service's API key and URL
    TEXT_TO_SPEECH_URL         # these two variables are required if CABOT_USE_ROBOT_TTS is 1
    
  • Options for simulation

    CABOT_INITX          # initial robot position x for gazebo (default=0)
    CABOT_INITY          # initial robot position y for gazebo (default=0)
    CABOT_INITZ          # initial robot position z for gazebo (default=0)
    CABOT_INITA          # initial robot angle (degree) for gazebo (default=0)
    
  • Others

    ## The following will be managed by docker-compose files
    ## be careful to set these variables in your .env file
    ##
    CABOT_GAZEBO             # 1: gazebo 0: real robot
    CABOT_TOUCH_ENABLED      # true: enabled - false: disabled
                               disabled for gazebo and enabled for real robot in docker-compose file
    CABOT_USE_REALSENSE      # to use realsense camera (default=0)
                               disabled for gazebo and enabled for real robot in docker-compose file
    CABOT_PRESSURE_AVAILABLE # to use pressure sensor (default=0)
                               disabled for gazebo and enabled for real robot in docker-compose file
    
  • DDS related

    NET_CORE_BUFFER_SIZE_IN_MB   # default 64, small buffer size can cause data congestion and degrade performance especially velodyne_points and localization
    

Navigate CaBot

  • Nav2 Goal tool does not work properly: the robot will move with the nav2 default BT xml (only for debugging purposes)
  • right-click on a blue dot in demo_2d_floors.rviz (-Y option for ros1 service to show) and select the "Navigate to Here" menu
  • or directory publish a /cabot/event topic on ROS1. see here more detail about destinations.
    # example destination in cabot_site_cmu_3d environment
    $ rostopic pub -1 /cabot/event std_msgs/String "data: 'navigation;destination;EDITOR_node_1496171299873'"
    

CaBot app for iOS

TBD

Customization (build your own map)

See customization for more details.

Development Detail

See development for more details.

Getting Involved

Issues and Questions

Please use Issues for both issue tracking and your questions about the CaBot repository.

Developer Certificate of Origin (DCO)

The developer needs to add a Signed-off-by statement and thereby agrees to the DCO, which you can find below. You can add either -s or --sign-off to your usual git commit commands. If Signed-off-by is attached to the commit message, it is regarded as agreed to the Developer's Certificate of Origin 1.1.

https://developercertificate.org/

Developer's Certificate of Origin 1.1

By making a contribution to this project, I certify that:

(a) The contribution was created in whole or in part by me and I
    have the right to submit it under the open source license
    indicated in the file; or

(b) The contribution is based upon previous work that, to the
    best of my knowledge, is covered under an appropriate open
    source license and I have the right under that license to
    submit that work with modifications, whether created in whole
    or in part by me, under the same open source license (unless
    I am permitted to submit under a different license), as
    Indicated in the file; or

(c) The contribution was provided directly to me by some other
    person who certified (a), (b) or (c) and I have not modified
    it.

(d) I understand and agree that this project and the contribution
    are public and that a record of the contribution (including
    all personal information I submit with it, including my
    sign-off) is maintained indefinitely and may be redistributed
    consistent with this project or the open source license(s)
    involved.

License

MIT License


The following files/folder are under Apache-2.0 License

  • docker/bridge/ros1/nav2_msgs/

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.