add lab4

ROS_Core/src/Labs/Lab4/CMakeLists.txt

@@ -0,0 +1,123 @@
+cmake_minimum_required(VERSION 3.0.2)
+project(racecar_learning)
+
+
+## Find catkin macros and libraries
+## if COMPONENTS list like find_package(catkin REQUIRED COMPONENTS xyz)
+## is used, also find other catkin packages
+find_package(catkin REQUIRED COMPONENTS
+  dynamic_reconfigure
+  rospy 
+  std_msgs
+  nav_msgs
+  racecar_msgs
+)
+
+# Uncomment this if the package has a setup.py. This macro ensures
+# modules and global scripts declared therein get installed
+# See http://ros.org/doc/api/catkin/html/user_guide/setup_dot_py.html
+# catkin_python_setup()
+
+################################################
+## Declare ROS messages, services and actions ##
+################################################
+
+## To declare and build messages, services or actions from within this
+## package, follow these steps:
+## * Let MSG_DEP_SET be the set of packages whose message types you use in
+##   your messages/services/actions (e.g. std_msgs, actionlib_msgs, ...).
+## * In the file package.xml:
+##   * add a build_depend tag for "message_generation"
+##   * add a build_depend and a exec_depend tag for each package in MSG_DEP_SET
+##   * If MSG_DEP_SET isn't empty the following dependency has been pulled in
+##     but can be declared for certainty nonetheless:
+##     * add a exec_depend tag for "message_runtime"
+## * In this file (CMakeLists.txt):
+##   * add "message_generation" and every package in MSG_DEP_SET to
+##     find_package(catkin REQUIRED COMPONENTS ...)
+##   * add "message_runtime" and every package in MSG_DEP_SET to
+##     catkin_package(CATKIN_DEPENDS ...)
+##   * uncomment the add_*_files sections below as needed
+##     and list every .msg/.srv/.action file to be processed
+##   * uncomment the generate_messages entry below
+##   * add every package in MSG_DEP_SET to generate_messages(DEPENDENCIES ...)
+
+## Generate messages in the 'msg' folder
+# add_message_files(
+#   FILES
+#   Message1.msg
+#   Message2.msg
+# )
+
+## Generate services in the 'srv' folder
+# add_service_files(
+#   FILES
+#   Service1.srv
+#   Service2.srv
+# )
+
+## Generate actions in the 'action' folder
+# add_action_files(
+#   FILES
+#   Action1.action
+#   Action2.action
+# )
+
+## Generate added messages and services with any dependencies listed here
+# generate_messages(
+#   DEPENDENCIES
+#   std_msgs  # Or other packages containing msgs
+# )
+
+################################################
+## Declare ROS dynamic reconfigure parameters ##
+################################################
+
+## To declare and build dynamic reconfigure parameters within this
+## package, follow these steps:
+## * In the file package.xml:
+##   * add a build_depend and a exec_depend tag for "dynamic_reconfigure"
+## * In this file (CMakeLists.txt):
+##   * add "dynamic_reconfigure" to
+##     find_package(catkin REQUIRED COMPONENTS ...)
+##   * uncomment the "generate_dynamic_reconfigure_options" section below
+##     and list every .cfg file to be processed
+
+## Generate dynamic reconfigure parameters in the 'cfg' folder
+generate_dynamic_reconfigure_options(
+  cfg/controller.cfg
+)
+
+
+###################################
+## catkin specific configuration ##
+###################################
+## The catkin_package macro generates cmake config files for your package
+## Declare things to be passed to dependent projects
+## INCLUDE_DIRS: uncomment this if your package contains header files
+## LIBRARIES: libraries you create in this project that dependent projects also need
+## CATKIN_DEPENDS: catkin_packages dependent projects also need
+## DEPENDS: system dependencies of this project that dependent projects also need
+catkin_package(
+    CATKIN_DEPENDS 
+    rospy
+    std_msgs 
+    racecar_msgs 
+    nav_msgs 
+    dynamic_reconfigure
+)
+
+#############
+## Install ##
+#############
+
+# all install targets should use catkin DESTINATION variables
+# See http://ros.org/doc/api/catkin/html/adv_user_guide/variables.html
+
+## Mark executable scripts (Python etc.) for installation
+## in contrast to setup.py, you can choose the destination
+install(PROGRAMS
+  scripts/imitation_learning_node.py
+  DESTINATION ${CATKIN_PACKAGE_BIN_DESTINATION}
+)
+

ROS_Core/src/Labs/Lab4/cfg/controller.cfg

@@ -0,0 +1,11 @@
+#!/usr/bin/env python
+PACKAGE = "racecar_learning"
+
+from dynamic_reconfigure.parameter_generator_catkin import *
+
+gen = ParameterGenerator()
+
+gen.add("latency",  double_t,    0, "Latency compensation for control", 0.2, -1, 1)
+gen.add("ref_speed",  double_t,    0, "Reference speed for control", 1, 0, 1.5)
+
+exit(gen.generate("racecar_learning", "racecar_learning", "controller"))

ROS_Core/src/Labs/Lab4/launch/lab4.launch

@@ -0,0 +1,21 @@
+<launch>
+    <arg name="output"              default="screen"                doc="display output to screen or log file"/>
+    <arg name="odom_topic"          default="/SLAM/Pose"            doc="ROS topic for the pose"/>
+    <arg name="control_topic"       default="/Control"              doc="ROS topic for control input"/>
+    <arg name="path_topic"          default="/Routing/Path"         doc="ROS topic for reference path"/>
+    <arg name='model_path'          default=""                      doc='path to the model file'/>
+    <arg name='lr'                  default="0.01"                  doc='learning rate'/>
+
+
+    <node pkg="racecar_learning" type="learning_node.py" name="imitation_learning" output="$(arg output)">
+        <param name="odom_topic"           value="$(arg odom_topic)" />
+        <param name="control_topic"        value="$(arg control_topic)" />
+        <param name="path_topic"           value="$(arg path_topic)" />
+        <param name="package_path"         value="$(find racecar_learning)" />
+        <param name="simulation"           value="False" />
+        <param name="PWM_model"            value="$(find racecar_learning)/models/pwm.sav" />
+        <param name="model_path"           value="$(arg model_path)" />
+        <param name="lr"                   value="$(arg lr)" />
+    </node>
+
+</launch>

ROS_Core/src/Labs/Lab4/package.xml

@@ -0,0 +1,20 @@
+<?xml version="1.0"?>
+<package format="2">
+  <name>racecar_learning</name>
+  <version>0.1.0</version>
+  <description>The traj_tracking_ros package</description>
+
+  <maintainer email="[email protected]">Zixu Zhang</maintainer>
+  <license>BSD</license>
+
+  <author >Zixu Zhang</author>
+
+
+  <buildtool_depend>catkin</buildtool_depend>
+
+  <depend>rospy</depend>
+  <depend>std_msgs</depend>
+  <depend>racecar_msgs</depend>
+  <depend>dynamic_reconfigure</depend>
+  <depend>nav_msgs</depend>
+</package>

ROS_Core/src/Labs/Lab4/readme.md

@@ -0,0 +1,73 @@
+# Lab 4: Behavior Cloning
+In this lab, we will use a basic behavior cloning algorithm to train a model to drive the truck.
+
+
+**Make sure that your repo is up-to-dated**
+# Setup Environment
+## 1. Open a new terminal and SSH into your robot
+```
+ssh nvidia@<IP OF YOUR ROBOT>
+```
+## 2. Activate *ros_base* environment on your robot
+```
+conda activate ros_base
+```
+## 3. Install Pytorch
+```
+pip3 install torch
+```
+
+# Launch Learning Node
+## Open **three** new terminals and let's call them *T1*, *T2*, and *T3*
+### 1. In *T1*, SSH into your robot and launch SLAM
+```
+ssh nvidia@<IP OF YOUR ROBOT>
+cd ~/StartUp
+./start_ros.sh <IP OF YOUR ROBOT>
+```
+### 2. In *T2*, navigate to the repo on your PC
+```
+cd <REPO ON YOUR PC>/ROS_Core
+conda activate ros_base
+catkin_make
+source devel/setup.bash
+source network_ros_client.sh <IP OF YOUR ROBOT> <IP OF YOUR PC>
+roslaunch racecar_interface visualization.launch enable_routing:=false
+```
+### 3. Start the SLAM from RQT
+![](./asset/rqt_truck.png)
+### 4. In *T3*, SSH into your robot and start learning node
+```
+ssh nvidia@<IP OF YOUR ROBOT>
+cd <REPO ON YOUR ROBOT>/ROS_Core
+conda activate ros_base
+catkin_make
+source devel/setup.bash
+source network_ros_host.sh <IP OF YOUR ROBOT>
+roslaunch racecar_learning lab4.launch
+``` 
+
+# Start Training Online
+## 1. In the RQT, call the service *"learning/start_learn"* from RQT to start training. 
+## 2. In the RVIZ, use the *"2D Nav Goal"* to set a reference path for the robot
+A loop will be generated automatically as your robot's reference path, use your controller to drive the robot along the path.
+![](./asset/loop.png)
+## 3. In *T3*, you will see the loss be printed out.
+You can drive your robot along the reference path for a few loops, then simply let the robot stop on the track and wait the loss to converge.
+## 4. Once the loss converges, call the service *"learning/eval"* from RQT to pause the training and evalute the model.
+Hit the down button on your controller to start the evaluation. The robot will drive along the reference path.
+## 5. If the robot drives well, call the service *"learning/save_model"* from RQT to save the model and call the service *"learning/save_data"* to stop the training.
+Your model will be saved in folder ["ROS_Core/Labs/Lab4/models"](./models) on your robot, and the training data will be saved in folder ["ROS_Core/Labs/Lab4/data"](./data) on your robot.
+## 6. If you do not like your model, call the service *"learning/start_learn"* again from RQT to resume training again.
+
+# Traning Offline
+With data collected from the previous step, we can train the model offline using provided [iPython notebook](./scripts/offline_train.ipynb). You can train this on your own PC, which should be significantly faster than the computer on the robot.
+
+# Test the Model
+You can evaluate the model trained offline by using additional parameter during the launch of the learning node. In *T3*, relaunch the node using 
+```
+roslaunch racecar_learning lab4.launch model_path:=<PATH TO YOUR MODEL>
+```
+
+# Task: Train your own behavior cloning policy, test it on the robot and show it to your AIs
+