MoveIt framework overview in ROS - Step-by-Step Execution

Concept Flow - MoveIt framework overview

Start: Robot Description

↓

Planning Scene Setup

↓

Motion Planning Request

↓

Planning Pipeline

↓

Trajectory Generation

↓

Execution on Robot

↓

Feedback & Monitoring

↓

End

MoveIt starts with robot info, sets up the scene, plans motion, generates trajectory, executes it, and monitors feedback.

Execution Sample

ROS

robot = load_robot_description()
scene = setup_planning_scene(robot)
plan = plan_motion(scene, goal)
trajectory = generate_trajectory(plan)
execute_trajectory(trajectory)
monitor_feedback()

This code shows the main steps MoveIt takes to plan and execute robot motion.

Execution Table

Step	Action	Input	Output	Notes
1	Load robot description	URDF file	Robot model loaded	Robot shape and joints defined
2	Setup planning scene	Robot model + environment	Planning scene ready	Includes obstacles and robot state
3	Plan motion	Planning scene + goal	Motion plan found	Path from start to goal computed
4	Generate trajectory	Motion plan	Trajectory points	Smooth path for robot joints
5	Execute trajectory	Trajectory points	Robot moves	Robot follows planned path
6	Monitor feedback	Robot sensors	Status updates	Detect success or errors
7	End	N/A	N/A	Process complete

💡 All steps complete, robot reached goal or error handled

Variable Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	After Step 5	After Step 6	Final
robot	None	Loaded robot model	Loaded robot model	Loaded robot model	Loaded robot model	Loaded robot model	Loaded robot model	Loaded robot model
scene	None	None	Planning scene ready	Planning scene ready	Planning scene ready	Planning scene ready	Planning scene ready	Planning scene ready
plan	None	None	None	Motion plan found	Motion plan found	Motion plan found	Motion plan found	Motion plan found
trajectory	None	None	None	None	Trajectory points	Trajectory points	Trajectory points	Trajectory points
execution_status	None	None	None	None	None	Robot moves	Status updates	Complete or error

Key Moments - 3 Insights

Why do we need a planning scene before motion planning?

The planning scene includes the robot and obstacles. Without it, the planner can't avoid collisions. See step 2 and 3 in the execution_table.

What is the difference between a motion plan and a trajectory?

A motion plan is a path in space; a trajectory adds timing and smooth joint movements. Steps 3 and 4 show this difference.

How does MoveIt know if the robot reached the goal?

It monitors feedback from sensors after execution (step 6) to confirm success or detect errors.

Visual Quiz - 3 Questions

Look at the execution_table, what output is produced after step 3?

A Trajectory points

B Motion plan found

C Robot moves

D Planning scene ready

💡 Hint

Check the 'Output' column for step 3 in the execution_table.

At which step does the robot actually start moving?

A Step 5

B Step 4

C Step 6

D Step 2

💡 Hint

Look for 'Robot moves' in the 'Output' column of the execution_table.

If the planning scene is not set up correctly, which step will likely fail?

A Step 1

B Step 5

C Step 3

D Step 6

💡 Hint

Planning depends on the planning scene; see step 3 in the execution_table.

Concept Snapshot

MoveIt Framework Overview:
1. Load robot description (URDF).
2. Setup planning scene with robot and obstacles.
3. Plan motion path to goal.
4. Generate smooth trajectory.
5. Execute trajectory on robot.
6. Monitor feedback for success or errors.

Full Transcript

MoveIt framework helps robots move safely and smoothly. It starts by loading the robot's description, then sets up a planning scene including obstacles. Next, it plans a path from the current position to the goal. This plan is turned into a smooth trajectory that the robot can follow. The robot then executes this trajectory. Finally, MoveIt monitors sensors to check if the movement succeeded or if there were problems. Each step builds on the previous one to ensure safe and accurate robot motion.

Practice

1. What is the main purpose of the MoveIt framework in ROS?

easy

A. To help robots plan and move their arms safely

B. To control robot sensors and cameras

C. To manage robot battery and power levels

D. To handle robot communication protocols

Solution

1. Step 1: Understand MoveIt's role in ROS

   MoveIt is designed to help robots plan and execute arm movements safely and efficiently.

2. Step 2: Compare options with MoveIt's purpose

   The other options relate to sensors, power, and communication, which are not MoveIt's focus.

3. Final Answer:

   To help robots plan and move their arms safely -> Option A

4. Quick Check:

   MoveIt = arm planning and safe movement [OK]

Hint: Remember MoveIt is about arm movement planning [OK]

Common Mistakes:

• Confusing MoveIt with sensor or power management
• Thinking MoveIt handles communication protocols
• Assuming MoveIt controls the whole robot

2. Which of the following is the correct way to define a MoveIt planning group named "gripper" in a ROS configuration?

easy

A.

B.

C.

D. arm

Solution

1. Step 1: Identify correct XML tag for MoveIt groups

   MoveIt uses the <group> tag with a name attribute to define planning groups such as 'gripper'.

2. Step 2: Check options for valid syntax

   <group name="gripper" /> correctly uses the proper syntax and name. <group name="arm" /> has the correct tag format but wrong name value. <moveit_group name="arm" /> and <planning_group>arm</planning_group> use incorrect tags and structure.

3. Final Answer:

   <group name="gripper" /> -> Option B

4. Quick Check:

   MoveIt groups use <group name="..." /> [OK]

Hint: Look for syntax in MoveIt configs [OK]

Common Mistakes:

• Using incorrect XML tags like <planning_group>
• Confusing group names with tags
• Missing the name attribute in group tag

3. Given the following MoveIt code snippet, what will be the output if the goal pose is reachable?

move_group.setPoseTarget(goal_pose);
bool success = (move_group.plan(my_plan) == moveit::planning_interface::MoveItErrorCode::SUCCESS);
std::cout << "Planning success: " << success << std::endl;

medium

A. Planning success: false

B. Planning success: 0

C. Planning success: 1

D. Planning success: true

Solution

1. Step 1: Understand the return type of plan()

   The plan() function returns a MoveItErrorCode. Comparing it to SUCCESS returns true or false, which converts to 1 or 0 when printed as an integer.

2. Step 2: Check output of std::cout for bool

   std::cout prints bool as 1 for true and 0 for false by default, so if the goal is reachable, success is true, printed as 1.

3. Final Answer:

   Planning success: 1 -> Option C

4. Quick Check:

   Bool true prints as 1 in std::cout [OK]

Hint: Bool prints as 1 or 0 in C++ std::cout [OK]

Common Mistakes:

• Expecting 'true' or 'false' strings instead of 1 or 0
• Confusing success variable type
• Assuming plan() returns a boolean directly

4. You wrote this MoveIt code but the robot arm does not move:

move_group.setPoseTarget(goal_pose);
move_group.move();

What is the most likely reason?

medium

A. You forgot to call move_group.plan() before move()

B. move() does not execute the planned motion

C. You need to call move_group.stop() before move()

D. The goal_pose is not set correctly or unreachable

Solution

1. Step 1: Understand move_group.move() behavior

   move() plans and executes the motion internally, so calling plan() first is not required.

2. Step 2: Check common reasons for no movement

   If the goal_pose is invalid or unreachable, the robot will not move despite calling move().

3. Final Answer:

   The goal_pose is not set correctly or unreachable -> Option D

4. Quick Check:

   Invalid goal_pose stops movement [OK]

Hint: Check if goal pose is valid and reachable first [OK]

Common Mistakes:

• Thinking plan() must be called before move()
• Assuming move() only executes without planning
• Calling stop() unnecessarily before move()

5. You want to control both the robot arm and gripper using MoveIt. How should you organize your MoveIt configuration to plan motions for both parts together?

hard

A. Create a single planning group including both arm and gripper joints

B. Use separate planning groups and plan motions independently

C. Control the gripper manually outside MoveIt and plan only the arm

D. Use a custom plugin to merge arm and gripper groups at runtime

Solution

1. Step 1: Understand MoveIt planning groups

   Planning groups define which joints MoveIt controls together. To move arm and gripper simultaneously, they must be in one group.

2. Step 2: Compare options for combined control

   Create a single planning group including both arm and gripper joints enables coordinated planning. Using separate planning groups and planning motions independently or controlling the gripper manually outside MoveIt may cause uncoordinated motions. Using a custom plugin to merge arm and gripper groups at runtime is complex and unnecessary for typical use.

3. Final Answer:

   Create a single planning group including both arm and gripper joints -> Option A

4. Quick Check:

   One group controls multiple parts together [OK]

Hint: Combine joints in one group for coordinated moves [OK]

Common Mistakes:

• Planning arm and gripper separately causing conflicts
• Trying manual control instead of MoveIt planning
• Overcomplicating with custom plugins unnecessarily