Developing complex robotics applications from scratch can be a tedious task. Even if we develop it, if our software is not modular and not reusable by others, the same process must be repeated by others, which is like reinventing the wheel. This situation applies not only to robotics but also to other technology domains. That’s why we come across a lot of software frameworks and libraries. These frameworks have ready-made functions or an Application Programming Interface (API) of commonly used algorithms that can be integrated into our program to help us quickly develop our applications.

In robotics software development, we have several software frameworks to create robot applications quickly. One of the popular frameworks available is ROS. Using ROS, we can quickly develop robotics applications by reusing existing code that other developers have already developed. Before jumping into the details, we will start by addressing the...

In this section, we will examine the fundamental architectural changes of ROS 1 and ROS 2 and the important reasons why the ROS 2 project started. Figure 1.3 shows the architecture of ROS 1 and ROS 2.

Figure 1.3: ROS 1 and ROS 2 architecture comparison

As you can see in Figure 1.3, there are three main layers in both of the architectures. They are:

• OS Layer
• Middleware Layer
• Application Layer

Let’s take a look at each layer in detail.

OS layer

On the left, you can see the architecture of ROS 1, and on the right, ROS 2. The bottom layer is the OS layer. As you know, ROS is a software framework that can be installed on an OS. ROS 1 is mainly compatible with Ubuntu/Linux, whereas ROS 2 is compatible with Linux, Windows, macOS, RTOS, etc.

Middleware layer

The next layer is the middleware layer, which constitutes the main ROS components. If you look at the ROS 1 architecture, it has...

In this section, we will discuss the important reasons for migrating to ROS 2 from ROS 1. Here, we will not discuss the technical aspects of it but, more generally, why it is better to migrate to ROS 2.

Here are the important points to note:

• ROS 1 Noetic Ninjemys will reach EOL in 2025. After EOL, we will not have new features, bug fixes, or security updates for the ROS 1 packages. There will not be any ROS 1 versions after this; ROS 1 Noetic is the final official distribution of ROS 1. The primary OS supporting ROS 1 Noetic is Ubuntu 20.04, which is also going to EOL in 2025, so there won’t be any OS updates as well. After 2025, we will only have ROS 2 distributions.
• ROS 2 can be used in any robotic application, including research and commercial applications. ROS 1 is mainly designed for academic and research use. If you plan to prototype and commercialize a robot, ROS 2 is a good choice because of its enhanced...

In this section, we discuss more details about ROS 2’s core module, DDS, which works as the communication middleware for ROS 2, as shown in Figure 1.3.

What is DDS?

DDS [21] is a standard from the Object Management Group (OMG) [22] that was established in 2004 for standard publisher/subscriber communication. It is mainly used in mission-critical applications like air traffic control and management, financial trading, and complex telemetry systems. DDS enables scalable, real-time, dependable, and high-performance data exchange between the publisher and subscriber.

The OMG is a global, open-membership, non-profit consortium responsible for creating and maintaining a number of standards in technical and business domains.

Publisher/subscriber communication involves sending and receiving data. The publisher sends data, and the subscriber listens to and receives it.

The DDS standard mainly came into existence when a group of companies...

We have gone through the basics of DDS and explored its different features. In this section, we are going to explore the layers in ROS 2 that work on top of DDS. We can start with a detailed discussion of the RMW abstraction layer, which is one of the important layers connecting DDS to the ROS 2 framework.

RMW layer (ROS middleware abstraction interface)

Figure 1.8 shows a detailed architecture of the ROS 2 framework with all its layers.

Figure 1.8: ROS 2 detailed architecture (source: RTI | https://www.slideshare.net/slideshow/deep-dive-into-the-opc-ua-dds-gateway-specification/99648467)

We already discussed the DDS framework in Figure 1.2. We learned that DDS has an application layer or APIs that we can use to build our own apps. This is where ROS becomes relevant. The ROS 2 framework is built on top of the DDS API layer. If you take a look at Figure 1.8, you can see that after the DDS framework layer comes the ROS 2 framework...

This chapter introduced ROS 2, a very popular framework for the development of robotic applications. The chapter provided an overview of the role played by ROS in the provision of modularity and reusability in the platform, allowing developers to quickly create and adapt robotic applications through the reutilization of existing code. It provided an overview of key elements of ROS, including inter-process communication infrastructure, hardware abstraction, licensing under open-source terms, and huge libraries and tools for tasks like visualization, debugging, and communication between different parts of a robot.

It also drew some comparisons between ROS 1 and ROS 2. The latter is more robust because of DDS, has enhanced security features, and is cross-platform compatible. It provides insight into the architectural improvements in ROS 2, focusing on the use of DDS as a communication middleware that can offer flexibility and high performance in real-time applications. The...

• [1] https://www.ros.org/
• [2] https://www.openrobotics.org/
• [3] https://www.openrobotics.org/
• [4] https://roscon.ros.org/
• [5] https://www.openrobotics.org/
• [6] https://github.com/ros2/rviz
• [7] https://github.com/ros-visualization/rqt
• [8] https://nav2.org/
• [9] https://moveit.picknik.ai/
• [10] https://opencv.org/
• [11] https://pointclouds.org/
• [12] https://www.open3d.org/
• [13] https://www.ros.org/blog/community/
• [14] https://discourse.ros.org/
• [15] https://robotics.stackexchange.com/
• [16] https://www.ros.org/blog/discord/
• [17] https://www.linkedin.com/company/willow-garage/
• [18] https://wiki.ros.org/Distributions
• [19] https://docs.ros.org/en/jazzy/Releases.html
• [20] https://docs.ros.org/en/jazzy/Installation/DDS-Implementations.html
• [21] https://www.dds-foundation.org/
• [22] https://www.omg.org/
• [23] https://www.cloudflare.com/en-gb/learning...