A quick Recap on why ROS 2 is invented to improve upon ROS.

I Why ROS 2

Feature	ROS	ROS 2
Computational Resources	Strong, local resources	Potentially limited locally
Number of Agents	Few agents	Hundreds of agents
Network Reliability	Stable and reliable	Not always reliable
Application	Research-oriented	Commercial and industrial use

1. Limitations of ROS:

ROS is no longer used exclusively for scientific research.
More ROS-based products are being commercialized.
The increasing variety of application scenarios introduces additional requirements for the system.

2. Various demands in the era of intelligent robotics:

Streamlined methods for multi-robot systems
heterogeneous system compatibility
Real-time performance capabilities
Robustness to changing network conditions
Better suitability for commercial products
Lifetime project management

II Main differences

1. A revolution on architecture

In ROS, all nodes are managed by a single node called “Master”. This increases the risk of single point failure. In ROS2, the discovery mechanism based on DDS mitigates the risk.

Date Distribution Service (DDS): A middleware standard that enables real-time, decentralized communication between applications.

The architectures of ROS and ROS 2.

2. Re-design of the API

ROS 2 adopts the latest C++ standards and Python 3 while retaining most of the familiar usage patterns.

3. Build-process Upgrade

In ROS 2, the build process is based on ament and colcon, replacing the older rosbuild and catkin systems used in ROS 1. Ament provides a more modular and extensible build framework, while colcon acts as a higher-level build tool that supports parallel builds, better dependency handling, and cleaner workspace management. Together, they offer a more reliable and scalable workflow, especially for large, multi-package ROS projects.

ROS 2 Essentials