fix: update project index.md by ypilseong

src/cobots2024/book/projects/202121010/index.md

@@ -51,3 +51,256 @@
 
 기술적 실패 및 일정 지연에 대비하여 소프트웨어 및 하드웨어에 대한 충분한 이해와 테스트를 통해 위험 최소화
 
+## 프로젝트 결과
+
+## 목차 
+
+1. [실행 방법](#실행-방법)
+2. [주요 코드](#주요-코드)
+3. [실행 결과](#실행-결과)
+4. [한계점](#한계점)
+
+```{notice}
+이번 ROS 프로젝트는 Docker에서 진행을 하며python을 통해 구현을 했습니다.
+```
+
+## 실행 방법
+
+1. Docker image
+    - Docker image는 다음과 같습니다:
+        `ghcr.io/ros-planning/moveit2_tutorials:$DOCKER_IMAGE`
+    - Docker compose
+        ```yml
+        # Example command:
+        # Humble on a Nvidia system:
+        # DOCKER_IMAGE=humble-tutorial docker compose run gpu
+        # Rolling without discrete graphics:
+        # DOCKER_IMAGE=rolling-tutorial docker compose run cpu
+
+        services:
+        cpu:
+            image: ghcr.io/ros-planning/moveit2_tutorials:$DOCKER_IMAGE
+            container_name: moveit2_container
+            privileged: true
+            network_mode: host
+            command: /bin/bash
+            volumes:
+            - /tmp/.X11-unix:/tmp/.X11-unix
+            - $XAUTHORITY:/root/.Xauthority
+            - $PWD:/root/ws_moveit/src/moveit2_tutorials
+            environment:
+            QT_X11_NO_MITSHM: 1
+            DISPLAY: $DISPLAY
+        gpu:
+            image: ghcr.io/ros-planning/moveit2_tutorials:$DOCKER_IMAGE
+            container_name: moveit2_container
+            privileged: true
+            network_mode: host
+            command: /bin/bash
+            deploy:
+            resources:
+                reservations:
+                devices:
+                    - driver: nvidia
+                    count: 1
+                    capabilities: [gpu]
+            volumes:
+            - /tmp/.X11-unix:/tmp/.X11-unix
+            - $XAUTHORITY:/root/.Xauthority
+            - $PWD:/root/ws_moveit/src/moveit2_tutorials
+            environment:
+            QT_X11_NO_MITSHM: 1
+            DISPLAY: $DISPLAY
+            NVIDIA_VISIBLE_DEVICES: all
+            NVIDIA_DRIVER_CAPABILITIES: all
+        ```
+
+    - Docker 내부
+        ```sh
+        export DOCKER_IMAGE=humble-tutorial
+        export XAUTHORITY=~/.Xauthority
+
+        docker compose -f .docker/docker-compose.yml run gpu
+        ```
+
+2. Docker 내부에서 launch 파일 실행.
+    - Jupyter를 통해 구현했기 때문에 jupyer lunch.py 실행
+    ```sh
+    ros2 launch moveit2_tutorials jupyter_notebook_prototyping.launch.py start_servo:=true
+    ```
+
+3. ros2_project.ipynb 파일 실행
+
+## 주요 코드
+
+### 필요한 라이브러리 임포트 및 moveit config 작성
+
+```python
+import os
+import sys
+import yaml
+import rclpy
+import random
+import numpy as np
+from geometry_msgs.msg import PoseStamped, Pose
+from moveit.planning import MoveItPy
+from moveit.core.robot_state import RobotState
+from moveit_configs_utils import MoveItConfigsBuilder
+from ament_index_python.packages import get_package_share_directory
+from rclpy.node import Node
+from trajectory_msgs.msg import JointTrajectory, JointTrajectoryPoint
+
+# we need to specify our moveit_py config at the top of each notebook we use. 
+# this is since we will start spinning a moveit_py node within this notebook.
+
+moveit_config = (
+        MoveItConfigsBuilder(robot_name="panda", package_name="moveit_resources_panda_moveit_config")
+        .robot_description(file_path="config/panda.urdf.xacro")
+        .trajectory_execution(file_path="config/gripper_moveit_controllers.yaml")
+        .moveit_cpp(
+            file_path=os.path.join(
+                get_package_share_directory("moveit2_tutorials"),
+                "config",
+                "jupyter_notebook_prototyping.yaml"
+        )
+    )
+    .to_moveit_configs()
+    ).to_dict()
+```
+
+### rcply 초기화 및 panda 불러오기
+
+```python
+# initialise rclpy (only for logging purposes)
+rclpy.init()
+
+# instantiate moveit_py instance and a planning component for the panda_arm
+panda = MoveItPy(node_name="moveit_py", config_dict=moveit_config)
+panda_arm = panda.get_planning_component("panda_arm")
+```
+
+### 기능 코드
+
+```python
+def plan_and_execute(
+    robot,
+    planning_component,
+    single_plan_parameters=None,
+    multi_plan_parameters=None,
+):
+    """A helper function to plan and execute a motion."""
+    # plan to goal
+    if multi_plan_parameters is not None:
+        plan_result = planning_component.plan(
+            multi_plan_parameters=multi_plan_parameters
+        )
+    elif single_plan_parameters is not None:
+        plan_result = planning_component.plan(
+            single_plan_parameters=single_plan_parameters
+        )
+    else:
+        plan_result = planning_component.plan()
+
+    # execute the plan
+    if plan_result:
+        robot_trajectory = plan_result.trajectory
+        robot.execute(robot_trajectory, controllers=[])
+    else:
+        print("Planning failed")
+        
+
+
+# 특정 좌표로 이동시키는 함수
+def move_to_pose(robot, planning_component, pose):
+    planning_component.set_start_state_to_current_state()
+    planning_component.set_goal_state(pose_stamped_msg=pose, pose_link="panda_link8")
+    plan_and_execute(robot, planning_component)
+
+# 그리퍼를 여는 함수 (가상의 함수로 대체)
+def open_gripper():
+    print("Gripper opened")
+
+# 그리퍼를 닫는 함수 (가상의 함수로 대체)
+def close_gripper():
+    print("Gripper closed")
+   
+    
+place_pose_xy = [(-0.2, -0.2), (-0.25, -0.2),(-0.2, -0.25),(-0.25,-0.25),(-0.2,-0.3),(-0.25,-0.3)]
+
+# 초기 위치 정의 (특정 좌표로 설정)
+initial_pose = PoseStamped()
+initial_pose.header.frame_id = "panda_link0"
+initial_pose.pose.orientation.w = 0.0
+initial_pose.pose.orientation.x = 0.0
+initial_pose.pose.orientation.y = 1.0
+initial_pose.pose.orientation.z = 0.0
+initial_pose.pose.position.x = 0.3
+initial_pose.pose.position.y = 0.3
+initial_pose.pose.position.z = 0.7
+
+# 목표 좌표 정의
+pick_pose = PoseStamped()
+pick_pose.header.frame_id = "panda_link0"
+pick_pose.pose.orientation.w = 0.0
+pick_pose.pose.orientation.z = 0.0
+pick_pose.pose.orientation.y = 1.0
+pick_pose.pose.orientation.z = 0.0
+#pick_pose.pose.position.x = -0.28
+#pick_pose.pose.position.y = -0.2
+pick_pose.pose.position.z = 0.1
+
+place_pose = PoseStamped()
+place_pose.header.frame_id = "panda_link0"
+place_pose.pose.orientation.w = 0.0
+place_pose.pose.orientation.x = 0.0
+place_pose.pose.orientation.y = 1.0
+place_pose.pose.orientation.z = 0.0
+#place_pose.pose.position.x = 0.5
+#place_pose.pose.position.y = 0.0
+place_pose.pose.position.z = 0.3
+
+
+for i in range(6):
+    # 1. 초기 위치로 이동
+    move_to_pose(panda, panda_arm, initial_pose)
+
+    # 2. 특정 위치로 이동
+    x = random.uniform(0.0, 0.5)  # Random x position
+    y = random.uniform(0.0, 0.5)  # Random y position
+
+    pick_pose.pose.position.x = x
+    pick_pose.pose.position.y = y
+    move_to_pose(panda, panda_arm, pick_pose)
+
+    # 3. gripper 통해 객체 잡기 (구현x)
+    open_gripper()
+
+    close_gripper()
+
+    x, y = place_pose_xy[i]
+    place_pose.pose.position.x = x
+    place_pose.pose.position.y = y
+    
+    # 4. 객체를 특정 위치로 이동
+    move_to_pose(panda, panda_arm, place_pose)
+    
+    # 5. 객체 놓기
+    open_gripper()
+
+    move_to_pose(panda, panda_arm, initial_pose)
+```
+
+## 실행 결과
+
+실행 결과는 다음 영상과 같습니다.
+
+<video width="700px" controls="true" autoplay="true" loop="true">
+    <source src="https://github.com/ypilseong/cobots-2024/assets/104025187/829a26c1-de79-4ef9-851c-9901c34a9ac7" type="video/mp4">
+</video>
+
+## 한계점
+
+1. MoveIt2의 장애물 객체 지원 부족:
+    - 현재 MoveIt2는 Python 인터페이스를 통해 장애물 객체를 생성하는 기능을 지원하지 않습니다. 이로 인해 시뮬레이션 환경에서 장애물 객체를 생성하지 못했습니다.
+2. Gripper 구현 문제
+    - Gripper에 대한 자료를 충분히 찾지 못해 해당 기능을 구현하지 못했습니다.