Concepts and Packages

Core Concepts of ROS 1

Tutorials: https://wiki.ros.org/ROS/Tutorials

ROS 1 is structured around several key concepts that enable modular, distributed, and scalable robot software development. The main ideas are organized at different levels25:

1. Nodes

Definition: Nodes are independent processes that perform computation. Each node is typically responsible for a single task (e.g., reading sensor data, controlling motors, localization).
Implementation: Nodes are written using ROS client libraries such as roscpp (C++) or rospy (Python).
Communication: Nodes communicate with each other using topics, services, or the parameter server, and are managed at runtime by the ROS Master23 5.

2. Master

Role: The ROS Master provides name registration and lookup for nodes, topics, and services, enabling nodes to discover and connect with each other.
Operation: Nodes register themselves and their communication interfaces with the Master, which acts like a DNS server for the ROS network. After discovery, nodes communicate directly25.

3. Parameter Server

Purpose: Stores global parameters (key-value pairs) accessible at runtime by all nodes. Used for configuration and tuning of nodes without code changes25.

4. Topics

Mechanism: Topics are named message buses for asynchronous, many-to-many communication. Nodes publish messages to topics, and other nodes subscribe to receive those messages.
Strong Typing: Each topic enforces a specific message type for data consistency.
Transport: Default is TCPROS (reliable TCP/IP), with UDPROS (UDP-based, low-latency) for specific use cases23 5.

5. Messages

Definition: Strictly defined data structures used to communicate over topics. Defined in .msg files and can include primitive types or nested messages3 5.

6. Services

Mechanism: Synchronous, request/response interactions for tasks that require confirmation or a result (e.g., requesting a sensor reading).
Definition: Defined in .srv files with request and response fields5.

7. Bags

Purpose: Files (with .bag extension) for recording and replaying topic data. Essential for debugging, simulation, and development5.

8. Filesystem Level

Packages: The basic unit of ROS software organization, containing nodes, libraries, configuration files, and more.
Workspaces: Directories for building and managing multiple packages together5.

Popular and Essential ROS 1 Packages

While the importance of a package depends on your application, the following are widely used and considered essential in most ROS 1 projects4 6 7:

Package Name

Purpose/Description

tf / tf2

Coordinate frame transforms (essential for robot kinematics)

rviz

3D visualization tool for sensor data, robot models, and planning

ros_control

Framework for controller management and hardware abstraction

moveit

Motion planning and manipulation for robotic arms

navigation

Autonomous navigation stack for mobile robots

gazebo_ros_pkgs

Integration between ROS and Gazebo simulator

urdf

Unified Robot Description Format for robot models

robot_localization

State estimation from multiple sensors

rosbag

Data recording and playback

sensor_msgs, geometry_msgs, std_msgs

Standard message types for sensors, geometry, etc.

image_transport, cv_bridge

Tools for handling images and computer vision

robot_state_publisher

Publishes robot joint states to tf

actionlib

Tools for asynchronous task execution

ros_perception

Perception stack for vision and point cloud processing

rosparam, rosnode, rostopic, rosservice

Core tools for managing ROS system

ROS Industrial

Packages for industrial robot support

rqt

GUI plugins for introspection, plotting, and debugging

Other Notable Packages (from usage statistics)6:

cv_bridge, image_transport (vision)
robot_state_publisher, joint_state_publisher (robot state)
dynamic_reconfigure (runtime parameter tuning)
catkin (build system)
roslaunch, rosbag, rosmsg, rostopic, rosservice, rosnode (core utilities)
urdf_tutorial, turtlesim (learning and demos)

References and Further Reading

ROS 2 Concepts and Popular Packages

ROS 2 Tutorials: https://docs.ros.org/en/humble/Tutorials.html

Core Concepts of ROS 2

ROS 2 is a modern, distributed robotics middleware designed for flexibility, scalability, and real-time performance. It builds on the lessons of ROS 1 and introduces several key architectural and functional improvements.

1. Nodes

Definition: Nodes are independent, modular processes that perform computation. Each node can handle a specific function, such as sensor reading, data processing, or actuator control.
Types:
- Standard Node: Basic node with minimal state management.
- Lifecycle Node: Supports managed states (e.g., unconfigured, inactive, active, finalized), enabling robust startup, shutdown, and error handling4 7.
Features: Nodes are more modular in ROS 2, can be composed into a single process, and support advanced lifecycle management for reliability4 7.

2. Topics

Mechanism: Topics are named channels for asynchronous, many-to-many publish/subscribe communication. Nodes publish messages to topics, and other nodes subscribe to receive them24 6.
DDS Integration: ROS 2 uses the Data Distribution Service (DDS) middleware, enabling automatic discovery, decentralized communication, and advanced Quality of Service (QoS) controls6 7.
QoS Policies: Fine-grained control over reliability, durability, deadline, and resource usage, crucial for real-time and safety-critical applications4 7.

3. Services

Definition: Services provide synchronous, request/response communication. A client sends a request to a service server and waits for a response, suitable for tasks that require confirmation or a result4 6.
Implementation: Defined using .srv files specifying request and response types.

4. Actions

Purpose: Actions allow for asynchronous, long-running tasks with feedback and the ability to cancel (e.g., navigation goals, manipulation tasks).
Structure: Built on top of topics and services, actions provide goal, feedback, and result channels4 6.

5. Parameters

Role: Store configuration and runtime values for nodes. Parameters can be dynamically updated, allowing for flexible system tuning4 6.

6. Quality of Service (QoS)

Importance: QoS policies (enabled by DDS) let developers specify message delivery guarantees, reliability, and resource constraints for each topic or service, enhancing real-time and distributed performance4 7.

7. Launch System

Advancement: ROS 2 uses Python-based launch files, allowing for programmable, flexible system startup and orchestration4 6.

8. Security

Features: Built-in support for authentication, encryption, and access control, making ROS 2 suitable for commercial and safety-critical deployments4 7.

9. Tools and Ecosystem

Developer Tools: ROS 2 offers a suite of tools for introspection, debugging, visualization (e.g., RViz2), plotting, logging, and playback, accelerating development and troubleshooting5 7.
Community: A vibrant, global community contributes to a rich repository of open-source packages, ensuring continuous improvement and innovation7.

Popular and Essential ROS 2 Packages

The following packages are widely used and form the backbone of most ROS 2 applications. Your choice may vary depending on your robot and application, but these are considered foundational:

Package Name

Purpose/Description

rclcpp / rclpy

Core C++ and Python client libraries for node development

tf2

Coordinate frame transformations for robot kinematics and sensor fusion

RViz2

3D visualization tool for sensor data, robot models, and planning

ros2_control

Modular, extensible robot control framework

Nav2

Navigation stack for autonomous mobile robots

MoveIt 2

Advanced motion planning and manipulation for robotic arms

Gazebo / Ignition

High-fidelity robot simulation environments

sensor_msgs, geometry_msgs, std_msgs

Standard message types for sensors, geometry, and basic data

image_tools, camera_info_manager

Vision and sensor integration tools

rosbag2

Data recording and playback for ROS 2 messages

robot_state_publisher

Publishes robot joint states to tf2

action_msgs, lifecycle_msgs

Standard messages for actions and node lifecycle management

rqt, RViz2 plugins

GUI tools for introspection, plotting, debugging, and visualization

slam_toolbox

SLAM (Simultaneous Localization and Mapping) for mapping and localization

navigation2 (Nav2)

Advanced navigation and path planning for mobile robots

Perception packages

Includes packages for vision, point cloud processing, and object detection

Other Notable Packages:

micro_ros (ROS 2 for microcontrollers)
ament_cmake (build system)
colcon (build tool)
ros2launch (launch management)
ros2param (parameter management)
ros2topic, ros2service, ros2bag (core CLI tools)

References and Further Reading

PreviousROS NextManual and Quick Setup

Last updated 5 months ago