Awesome

ADLINK NeuronBot

Abstract

The purpose of this pkg is to demonstrate two main features of ADLINK Neuron miniITX platform.

1. Computing Power:
   The host robot is able to follow human by integrating all outputs from SPENCER robot framework, intel "object analytics" pkg and laser based leg_tracker pkg. ADLINK Neuron illustates its computing power to run these detecting/tracking algorithms smoothly. (HOG+SVM, CNN, Template matching, EKF, MobileNetSSD, etc)
2. ROS2/DDS Capabilities:
   The host robot (for following) is publishing its own pose with respect to known map through ROS2/DDS layer while following a human. On the other hand, the client robot (random wandering), could avoid the host robot by receiving host's pose and replaning its path.

[Official Slides] https://github.com/Adlink-ROS/adlink_neuronbot/blob/master/document/ADLINK_NeuronBot_20180313.pdf
[Youtube Video] https://youtu.be/RC6XvTvTs9Y
[Youtube Video] https://youtu.be/qA4_Hmnd_tM

Developer & Team

HaoChih Lin
Alan Chen
Chester Tseng
Bill Wang
Erik Boasson
Ryan Chen

ADLINK Technology, Inc
Advanced Robotic Platform Group

License

Apache 2.0
Copyright 2018 ADLINK Technology, Inc.

Tutorial

System Prerequisite

[Packages]

• Realsense D400
  Source: https://github.com/intel-ros/realsense
  Notice: About RealSense SDK 2.0, we highly recommed binary version.
  Testing: $ roslaunch realsense2_camera demo_pointcloud.launch
  <br />
• Astra Pro (alternative choose for camera)
  Source: https://github.com/orbbec/ros_astra_camera
  Notice: If possible, please buy Astra instead of Astra Pro!
  <br />
• YDLidar
  Source: https://github.com/EAIBOT/ydlidar
  Notice: could be replaced by any type of lidar.
  <br />
• Navigation
  Binary: $ sudo apt-get install ros-kinetic-navigation*
  Source: https://github.com/ros-planning/navigation
  Notice: if "replan" mode of global planner is malfunctioned, please compile whole pkgs from source.
  Testing: $ roslaunch spencer_people_tracking_launch tracking_on_bagfile.launch
  <br />
• SPENCER
  Binary: https://github.com/spencer-project/spencer_people_tracking#installation-from-l-cas-package-repository
  Source: https://github.com/spencer-project/spencer_people_tracking#installation-from-source
  Notice: Unless you want to use HOG+SVM [CUDA required], we highly recommend binary version.
  <br />
• Intel object analytics
  Source: https://github.com/intel/ros_object_analytics
  Notice: Tested with NCSDK v1.12, ros_object_analytics should be "devel" branch
  Testing:
  ** movidius_ncs **
  $ roslaunch realsense2_camera demo_pointcloud.launch
  $ roslaunch movidius_ncs_launch ncs_camera.launch cnn_type:=mobilenetssd input_topic:=/camera/color/image_raw
  $ roslaunch movidius_ncs_launch ncs_stream_detection_example.launch camera_topic:="/camera/color/image_raw"
  ** object_analytics **
  $ roslaunch realsense2_camera demo_pointcloud.launch
  $ roslaunch object_analytics_launch analytics_movidius_ncs.launch input_points:=/camera/depth/color/points
  <br />
• leg_tracker
  Source: https://github.com/angusleigh/leg_tracker
  Notice: The kinetic branch only supports OpenCv 3.3 and higher ver.
  Testing: $ roslaunch leg_tracker demo_stationary_simple_environment.launch
  <br />
• Turtlebot2
  Binaty: $ sudo apt-get install ros-kinetic-turtlebot
  Source: https://github.com/turtlebot/turtlebot
  Notice: We highly recommed you to install binary version.
  <br />

Launching Steps

• Mapping & Time Synchronizing
• Host robot (for following)
  $ roslaunch adlink_neuronbot NeuronBot_Demo_Host_AIO.launch
  OR (script)
  $ ./PATH_TO_WORKSPACE/adlink_neuronbot/autostart/NeuronBot_Demo_Host_AutoStart.sh
• Client robot (for avoidance)
  $ roslaunch adlink_neuronbot NeuronBot_Demo_Client_AIO.launch
  OR (script)
  $ ./PATH_TO_WORKSPACE/adlink_neuronbot/autostart/NeuronBot_Demo_Client_AutoStart.sh

Known Issues

• Q: move_base replanning does not work
  A: update your move_base pkg to the latest one, or compile it from source