Best ROS Labs and Hands On Learning Environments

By RWT Robotics Editorial Team Checked against 8 sources Published

A ROS lab in late 2026 runs on ROS 2 Lyrical Luth or Jazzy Jalisco with Gazebo, and you can build the first version for free in Docker on a laptop you already own. ROS 1 and Gazebo Classic are both end-of-life, so a “ROS lab” tutorial written around Noetic, catkin or TurtleBot3 on Gazebo 11 will send you down a dead branch. This guide sets out a lab in three tiers: a free virtual lab, a home lab on real hardware, and a classroom lab. Each tier comes with dated support windows, real prices and working configs.

New lab distroLyrical or Jazzy
SimulatorGazebo Jetty or Harmonic
Free tierDocker on any PC
Starter robotTurtleBot 4 Lite
Last checkedNovember 14, 2026

Key takeaways

  • Standardise new labs on ROS 2 Lyrical Luth (LTS to May 2031) or Jazzy Jalisco (LTS to May 2029).
  • Kilted Kaiju support ends in late 2026, Humble ends in May 2027, and Noetic ended in May 2025.
  • Gazebo Classic reached end-of-life in January 2025. Use Gazebo Jetty with Lyrical, or Harmonic with Jazzy.
  • In Docker, mount your workspace from the host so your work survives when the container is removed.
  • In a classroom, give every bench its own ROS_DOMAIN_ID so robots on shared Wi-Fi don’t cross-talk.

Fact-checked

We checked these key facts against the sources named, on November 14, 2026:

  • Humble reaches end-of-life in May 2027, so any course you build on it today needs porting within a semester or two (source: endoflife.date)
  • For Jazzy, the TurtleBot 4 user manual lists image v2.0.0 on Ubuntu Server 24.04.5, paired with Create 3 firmware I.0.0 (source: turtlebot.github.io)
  • On Apple Silicon, the most reliable route is an Ubuntu arm64 virtual machine running Jazzy or Lyrical natively, since both ship arm64 builds (source: docs.ros.org)
  • The Open Source Robotics Foundation (OSRF, operating as Open Robotics) maintains the core, and the Open Source Robotics Alliance (OSRA), which OSRF set up in 2024, handles project governance (source: openrobotics.org)
  • The biggest break is simulation: Humble-era tutorials often use Gazebo Classic (source: docs.nav2.org)

Which “Ros Lab” are you looking for?

There are two different things called a Ros Lab. One is a chemistry research group. The other is the robotics setup this page covers. If you came looking for Professor Alexandra Ros’s lab at Arizona State University, you want her bioanalytical microfluidics group. It sits in ASU’s School of Molecular Sciences and is affiliated with the Biodesign Institute. The group is known for dielectrophoresis and for microfluidic sample delivery for X-ray free-electron laser (XFEL) crystallography. Search ASU’s Biodesign Institute directory for her name.

Did you know?

The best-known “Ros Lab” on Google has nothing to do with robots. ASU News reported that Alexandra Ros’s group built microfluidic sample delivery that cut protein-crystal sample waste by more than 60% at XFEL facilities.

Everyone else is probably here to set up a Robot Operating System lab, so the rest of this page covers that.

What a ROS lab is (and what ROS actually is)

A ROS lab is any environment where you can write, run and debug Robot Operating System software against a simulated or real robot. Three versions cover almost every case: a virtual lab (containers plus simulation), a home lab (one dev PC plus one robot) and a classroom lab (many identical seats and several robots sharing a network).

ROS itself is a set of open-source libraries, tools and conventions for building robot software. It handles the messaging between processes (nodes), hardware abstraction, visualisation in RViz2, recording and replay of data, and a large ecosystem of packages for navigation, manipulation and perception. In ROS 2 the messaging layer runs on DDS. According to Open Robotics, Fast DDS has been the default middleware for 11 years and is still the default in Lyrical Luth. If you want the background on why the field moved, our explainer on what ROS 2 is and why more robots are moving to it covers that history.

Is ROS free?

Yes. ROS is free and open source, mostly under the Apache 2.0 and BSD licences, so you can use it commercially. The Open Source Robotics Foundation (OSRF, operating as Open Robotics) maintains the core, and the Open Source Robotics Alliance (OSRA), which OSRF set up in 2024, handles project governance. The software itself costs nothing, so a ROS lab’s costs come from hardware, computers, networking, any cloud services and ongoing maintenance.

Is ROS an operating system?

No. Despite the name, ROS is middleware and an SDK that runs on top of a real operating system, usually Ubuntu Linux. Your ROS version is tied to your Ubuntu version, which is why picking a distro is the first decision in any lab plan.

Pick your distro first: the 2026 support-window table

For a new lab in November 2026, choose Lyrical Luth if you’re starting from scratch on Ubuntu 26.04, or Jazzy Jalisco if you have hardware or courseware that already targets Ubuntu 24.04. Keep Humble only for existing courses you can’t port yet. Don’t start anything on Kilted.

Distro Type Released End of life Use in a new lab?
ROS 2 Lyrical Luth LTS (5 years) May 22, 2026 May 2031 Yes, best for new sim-first labs
ROS 2 Jazzy Jalisco LTS May 2024 May 2029 Yes, best for TurtleBot 4 hardware labs
ROS 2 Kilted Kaiju Non-LTS May 2025 December 2026 (ROS homepage says November) No
ROS 2 Humble Hawksbill LTS May 2022 May 2027 Legacy courseware only
ROS 2 Iron Irwini Non-LTS May 2023 December 4, 2024 No, already EOL
ROS Noetic (ROS 1) Final ROS 1 May 2020 May 2025 No, already EOL

Lyrical Luth is the twelfth ROS 2 release. Its binaries went out for testing on April 30, 2026, and the distro froze on May 18 before general availability on May 22.

Did you know?

A newer ROS distro can expire before an older one. Per endoflife.date, Kilted Kaiju (May 2025) loses support at the end of 2026, while Humble Hawksbill from 2022 lasts until May 2027.

Lyrical vs Jazzy for a lab

Both are long-term support releases, so you aren’t choosing between “stable” and “experimental”. The real differences are the operating system and the ecosystem.

Factor Lyrical Luth Jazzy Jalisco
Tier 1 platform Ubuntu 26.04 Resolute Raccoon Ubuntu 24.04 Noble
Ubuntu 24.04 support Tier 3 only Tier 1
Paired Gazebo Jetty Harmonic
Support ends May 2031 May 2029
TurtleBot 4 images Not yet documented Yes (v2.0.0 on Ubuntu Server 24.04.5)
Third-party package coverage Still filling in Mature

Pick Lyrical when your lab is mostly simulation, you’re installing fresh Ubuntu 26.04 machines (released April 23, 2026, with standard support to 2031), and you want the longest runway. Pick Jazzy when you have TurtleBot 4s or any robot whose vendor ships Jazzy images. Pick Jazzy too if you depend on packages that haven’t been released for Lyrical yet. In May 2026, community reports noted that slam_toolbox and cartographer were still missing from Lyrical.

The counter-argument for Jazzy is simple: two fewer years of support. For a three-year course plan that rarely matters. For a research lab buying robots now, Jazzy’s mature ecosystem is worth more than the extra two years.

What about Humble courseware?

The common forum complaint goes something like: “Should I learn on Humble or Jazzy? All the tutorials are Humble.” Learn on Jazzy. The ROS 2 concepts, CLI tools and client libraries (rclcpp, rclpy) carry over almost unchanged, so a Humble tutorial works on Jazzy for most beginner material. The biggest break is simulation: Humble-era tutorials often use Gazebo Classic. Humble reaches end-of-life in May 2027, so any course you build on it today needs porting within a semester or two.

Is there a ROS 3?

No. Open Robotics hasn’t announced a ROS 3. ROS 2 keeps shipping a new distro every May, with an LTS every other year. Rolling is the continuously updated development distro that feeds the next one. Treat any “ROS 3” headline as speculation until it appears on docs.ros.org.

Engineer's note

Put every lab seat on one LTS (Lyrical or Jazzy) and one Gazebo LTS. Mixed distros across benches cause more classroom failures than anything else, because message definitions and launch files drift between releases.

Tier 1: a free virtual ROS lab in Docker

The cheapest working ROS lab is a container on the computer you already have. Docker’s official ROS 2 guide starts with docker run -it ros:humble, and you can swap in jazzy or rolling. The guide also warns that files are lost when the container is removed unless you mount a persistent workspace. That’s exactly why so many forum threads begin “my files disappeared when I stopped the container.”

Image choice matters. The Docker Official Image ros:<distro> is a lean, command-line base. For RViz2 and Gazebo GUIs, use OSRF’s osrf/ros:<distro>-desktop-full images. Open Robotics confirmed osrf/ros:lyrical-desktop-full for amd64 and arm64 alongside osrf/ros:jazzy-desktop-full. Check the tag list on Docker Hub before pinning a version.

Linux host

On a Linux host, the GUI comes out of the container through the host’s X11 socket. Create a project folder with a ws/src directory, then add this Dockerfile. It bakes in your extra packages, so they don’t disappear either:

FROM osrf/ros:jazzy-desktop-full
RUN apt-get update && apt-get install -y \
      ros-jazzy-ros-gz \
      python3-colcon-common-extensions \
    && rm -rf /var/lib/apt/lists/*
RUN echo "source /opt/ros/jazzy/setup.bash" >> /root/.bashrc
WORKDIR /root/ws

Then add a compose.yaml next to it:

services:
  ros:
    build: .
    container_name: roslab
    network_mode: host
    ipc: host
    environment:
      - DISPLAY=${DISPLAY}
      - ROS_DOMAIN_ID=${ROS_DOMAIN_ID:-0}
      - QT_X11_NO_MITSHM=1
    volumes:
      - ./ws:/root/ws
      - /tmp/.X11-unix:/tmp/.X11-unix:rw
    devices:
      - /dev/dri:/dev/dri
    command: sleep infinity

Quick steps: start the Docker ROS lab

  1. Run xhost +local: on the host to allow local GUI clients.
  2. Run docker compose up -d –build in the project folder.
  3. Open a shell with docker compose exec ros bash.
  4. Build your code with colcon build inside /root/ws.
  5. Stop with docker compose stop; your ws folder stays on the host.

Check the lab with a talker, RViz2 and an empty Gazebo world:

ros2 run demo_nodes_cpp talker &
ros2 topic echo /chatter --once
rviz2 &
ros2 launch ros_gz_sim gz_sim.launch.py gz_args:=empty.sdf

After you build, reload your environment with source install/setup.bash so ros2 run can find your packages. For a Lyrical lab, replace every jazzy with lyrical. When you’re done, run xhost -local: to close the display permission again.

Engineer's note

The ./ws:/root/ws volume line is what makes this lab persistent. Your source code and build output live on the host disk, so docker compose down deletes nothing you wrote. Anything you install with apt by hand inside the container is still lost, so move it into the Dockerfile.

If you prefer an IDE, the ROS 2 docs include a guide to setting up ROS 2 with VS Code and Docker using dev containers. It wraps the same idea, an image plus a mounted workspace, in a .devcontainer folder, so every student opens an identical environment from a Git repo.

Windows (WSL2) and macOS

On Windows 11, run Ubuntu 24.04 (or 26.04) in WSL2 and install ROS natively inside it, or run the Docker setup above from the WSL2 shell. WSLg already provides a display server, so DISPLAY and /tmp/.X11-unix exist inside WSL. That answers the common question about containerised ROS 2 on WSL2: “how do I get Gazebo and RViz2 to show up?” Run the container from inside the WSL distro rather than from PowerShell, and mount the WSLg socket as shown. Gazebo’s 3D rendering is the most fragile piece. If it shows a black window, update your GPU driver on the Windows side first.

macOS is harder. ROS 2 on macOS isn’t a Tier 1 platform, and Docker Desktop’s Linux VM has no native display to forward. On Apple Silicon, the most reliable route is an Ubuntu arm64 virtual machine running Jazzy or Lyrical natively, since both ship arm64 builds.

The other common beginner question is: “Is it better to dual-boot Ubuntu or just use Docker?” Start in Docker or WSL2. It costs nothing and teaches ROS concepts just as well. Move to a native Ubuntu install once you connect real hardware, because USB devices, GPU rendering and DDS multicast all behave better without a virtualisation layer in the way.

Why not the old ros-lab VNC repos

Several “ros-lab” projects on GitHub predate ROS 2’s maturity. The one ranking for this keyword is a small ROS 1 and TurtleBot3 setup built on a third-party VNC desktop image rather than the official ros:<distro> images, and its author notes that it can’t stop or save containers. It also points to ROS 1 wiki tutorials. The current references are the docs.ros.org tutorials for your distro and Docker’s own ROS 2 guide.

Simulation: Gazebo Jetty, not Gazebo Classic

Use Gazebo Jetty for Lyrical labs and Gazebo Harmonic for Jazzy labs. Gazebo Classic (Gazebo 11) reached end-of-life in January 2025, as The Robot Report covered, so it no longer gets fixes and doesn’t pair with any supported ROS 2 LTS.

Version Status Support until Ubuntu requirement Pairs with
Gazebo Jetty (gz-sim 10) LTS, current May 2031 (releases page) Ubuntu 24.04 Noble or later Lyrical Luth
Gazebo Harmonic (gz-sim 8) LTS May 2029 (releases page) Ubuntu 22.04 or 24.04 Jazzy Jalisco
Gazebo Classic 11 End of life Ended January 2025 Up to Ubuntu 22.04 ROS 1 Noetic, Humble-era tutorials

Jetty is the tenth major Gazebo release and its LTS window started in September 2025. Its launch announcement said support would end in September 2030, but the releases page now says May 2031. Similarly, Harmonic was first announced as supported until September 2028, while the releases page lists May 2029. Plan on the earlier date of each pair.

Simulation is also where a lab pays for itself. A sim-first workflow, where every node runs against a virtual robot before it touches hardware, is a small classroom version of how industry teams use digital twins in robotics.

Harmonic vs Jetty

The choice follows your ROS distro, not the simulator’s features. ros-lyrical-desktop pulls in Jetty on its own, and the Gazebo install docs map Jazzy to Harmonic through ROS vendor packages. Mixing a non-default pairing is possible but means building ros_gz yourself, which isn’t worth it in a teaching lab. One early rough edge: community reports from May 2026 flagged a mouse-wheel zoom bug in Jetty on Ubuntu 26.04.

How to spot an outdated tutorial

You can usually tell a tutorial’s era within a few lines.

Engineer's note

If a tutorial runs gazebo instead of gz sim, or mentions roscore, catkin_make, rosrun or ROS_MASTER_URI, it was written for ROS 1 or Gazebo Classic. Skip it or port it.

Tier 2: a home ROS lab on real hardware

A home lab is one native Ubuntu dev PC plus one robot. The TurtleBot 4 Lite is the default choice because it is supported, documented and built for ROS 2. A Raspberry Pi-based DIY robot costs less if you’re comfortable with wiring and drivers.

For the dev PC, install Ubuntu 24.04 for Jazzy or Ubuntu 26.04 for Lyrical directly on the disk, then follow the Debian package install for your distro on docs.ros.org. A discrete GPU helps Gazebo but isn’t required for 2D navigation work. RAM matters more than CPU once Gazebo, RViz2 and Nav2 run together.

TurtleBot 4 Lite vs Standard

Both TurtleBot 4 models share the same iRobot Create 3 base, Raspberry Pi 4B computer and RPLIDAR A1 lidar. The Lite is the one to buy for learning ROS 2. Clearpath Robotics itself recommends the Lite for ROS 2 fundamentals.

Spec TurtleBot 4 Lite TurtleBot 4 Standard
Base iRobot Create 3 iRobot Create 3
Computer Raspberry Pi 4B Raspberry Pi 4B
Lidar RPLIDAR A1 RPLIDAR A1
Camera Luxonis OAK-D Lite Luxonis OAK-D Pro (night vision, wide angle)
Extras None Sensor tower, UI board with display, buttons, LEDs, extra USB-C
Payload 9 kg default, 15 kg custom 9 kg default, 15 kg custom
Launch price (2022) $1,195 $1,850
Current price $1,465 at RobotShop US, tariffs included Ask distributors; EU resellers list about โ‚ฌ2,395 to โ‚ฌ2,480

The Standard is worth the extra money only if your projects need the stronger OAK-D Pro depth camera or the onboard UI, for example vision courses in dim rooms. For SLAM, Nav2 and basic perception, the Lite does the same work.

For Jazzy, the TurtleBot 4 user manual lists image v2.0.0 on Ubuntu Server 24.04.5, paired with Create 3 firmware I.0.0. The older G and H firmware lines don’t work with Jazzy. The I-series firmware also comes in separate Fast DDS and Cyclone DDS builds, so flash the one that matches your middleware. Under Jazzy, the Create 3 talks to the Pi over the internal wired link and shouldn’t join Wi-Fi itself.

Reality check

TurtleBot 4 documentation covers Humble and Jazzy. Lyrical images for it aren’t documented yet, so a TurtleBot 4 lab should run Jazzy for now.

Budget DIY robots

A Raspberry Pi 5 robot is the cheapest real-hardware ROS lab. If you want a ROS 2 Raspberry Pi 5 project running with minimal fuss, flash 64-bit Ubuntu Server 24.04, which supports the Pi 5. Then install ros-jazzy-ros-base from the ROS apt repository, because Jazzy on arm64 Noble is a Tier 1 platform. Keep RViz2 and Gazebo on your desktop, and let the Pi run only drivers and lightweight nodes. Put the Pi and the desktop on the same ROS_DOMAIN_ID and they will discover each other.

The trade-off is time. With a TurtleBot 4 you work on robot software from day one. With a DIY chassis you spend your first weeks on motor drivers, odometry and URDF files, which is great for a mechatronics course and a distraction for a software course. ROBOTIS’s TurtleBot3 sits between the two: it’s smaller and cheaper than a TurtleBot 4 and has ROS 2 documentation, though most of the internet’s TurtleBot3 material is ROS 1-era.

Tier 3: a classroom or university ROS lab

A classroom ROS lab works when every seat is identical, every bench is isolated on the network, and students prove their code in simulation before they borrow a robot. Most failures in teaching labs come from breaking one of those three rules.

Tier Hardware Approx. cost Best for Limitations
1: Virtual Existing PC, Docker or WSL2 $0 Self-learners, intro courses, pre-lab homework No real sensors; GPU rendering can be fragile in VMs
2: Home Native Ubuntu PC + TurtleBot 4 Lite About $1,465 for the robot Hobbyists, grad students, portfolio projects One robot; no multi-robot practice
3: Classroom Identical seats + one TurtleBot 4 Lite per bench About $11,720 for 8 Lites, before PCs University and college courses Network setup, maintenance and battery logistics

One golden image for every seat

Build one image and deploy it everywhere. That can be a Docker image from a Dockerfile in your course repo, a VS Code dev container, or a cloned Ubuntu disk image. Pin the distro, Gazebo version and every apt package in it. When a student says “it works on my machine,” every machine in the room should be the same machine. Rebuild once per term, never mid-semester.

Networking many robots in one room

ROS 2 has no master node: nodes find each other automatically over DDS multicast. That is convenient for one robot and chaotic for twelve. Give each bench a unique ROS_DOMAIN_ID. The ROS 2 docs recommend values from 0 to 101 to avoid port clashes on Linux:


export ROS_DOMAIN_ID=7 export ROS_AUTOMATIC_DISCOVERY_RANGE=LOCALHOST

ROS_AUTOMATIC_DISCOVERY_RANGE replaced the older ROS_LOCALHOST_ONLY variable starting with Iron, so ignore older guides that use the deprecated one. Campus Wi-Fi often blocks or rate-limits multicast. In that case, use the Fast DDS Discovery Server, which the TurtleBot 4 manual documents as an alternative to its Simple Discovery mode. A dedicated lab router you control is the cleanest setup of all.

Engineer's note

Write the bench’s ROS_DOMAIN_ID on a label stuck to both the robot and the PC. Most “my robot is moving on its own” incidents turn out to be two benches sharing a domain ID.

Courseware and curriculum

A common complaint is that ROS 2 tutorials are too complex for beginners. The fix is order, not easier content. A sequence that works:

  1. CLI tools and turtlesim from the docs.ros.org beginner tutorials for your distro.
  2. One Python package with a publisher and subscriber in rclpy.
  3. Launch files, parameters and RViz2 against a simulated robot in Gazebo.
  4. SLAM and Nav2 in simulation, using the TurtleBot 4 simulation packages.
  5. The same code on a physical TurtleBot 4, one bench at a time.

To answer the forum question “What environments do you use for training and courses?”: most working course setups combine a container or dev container for every seat with a shared simulated robot and a small pool of physical robots. For younger students who aren’t ready for Linux, a block-coding educational robotics system makes a better first step than ROS.

Migrating an old ROS 1 lab

Moving from Noetic, TurtleBot3 and Gazebo Classic to ROS 2, TurtleBot 4 and Jetty or Harmonic means rewriting build files, launch files and simulation plugins. Most of your robot knowledge transfers. Much of your code needs porting rather than a recompile.

ROS 1 / Classic era ROS 2 / modern Gazebo
roscore and ROS_MASTER_URI No master; DDS discovery plus ROS_DOMAIN_ID
catkin_make / catkin build colcon build
rosrun, roslaunch (XML) ros2 run, ros2 launch (Python, XML or YAML)
rospy, roscpp rclpy, rclcpp
gazebo, gazebo_ros_pkgs gz sim, ros_gz
Classic plugins such as libgazebo_ros_diff_drive.so Gazebo systems such as gz-sim-diff-drive-system
rosbag (.bag files) ros2 bag (MCAP by default)

Quick steps: port a ROS 1 lab

  1. Pick one target distro and one Gazebo LTS for all seats.
  2. Rebuild your environment as a Dockerfile or golden image.
  3. Port message and service packages first, then nodes.
  4. Rewrite launch files and Gazebo plugins for ros_gz.
  5. Replace TurtleBot3 Classic worlds with Jetty or Harmonic worlds.
  6. Retire every handout that mentions roscore or catkin.

URDF and SDF robot descriptions mostly carry over, which saves real time. Budget the most effort for the simulation layer, since Gazebo Classic plugins and world files are where old courses break hardest.

FAQ

What is the Ros Lab at ASU?

It’s Professor Alexandra Ros’s bioanalytical research group at Arizona State University, in the School of Molecular Sciences and affiliated with the Biodesign Institute. The group works on microfluidics, dielectrophoresis and sample delivery for XFEL protein crystallography. It isn’t connected to robotics.

What is ROS used for?

ROS is used to build robot software: drivers, sensor processing, mapping, navigation, manipulation and visualisation. It runs on research robots, warehouse robots, field robots and much of the teaching hardware sold to universities. Many of the robotics companies leading the industry build on it or ship ROS 2 interfaces.

Is ROS free?

Yes. ROS is open source under permissive licences, and you pay nothing to use it, including commercially.

Is ROS an operating system?

No. It’s middleware and a toolset that runs on an operating system, usually Ubuntu, with each ROS distro tied to specific Ubuntu releases.

Which ROS 2 version should I use in 2026?

Use Lyrical Luth for new, simulation-heavy labs on Ubuntu 26.04. Use Jazzy Jalisco for labs with TurtleBot 4s or other robots that ship Jazzy images. Use Humble only for courses you’re actively porting before its May 2027 end-of-life.

Is there a ROS 3?

No. No official Open Robotics source has announced ROS 3. ROS 2 continues with yearly releases.

Can I run a ROS lab on Windows or Mac?

Yes, with limits. On Windows 11, use WSL2 with Ubuntu, which shows GUIs through WSLg. On a Mac, run an Ubuntu virtual machine. Both work well for learning, but native Ubuntu is better once real hardware is involved.

How much does it cost to set up a ROS lab?

A virtual lab costs nothing beyond a computer you already own. A home lab with a TurtleBot 4 Lite costs about $1,465 for the robot at current US retail. A classroom with eight robot benches costs roughly $11,720 for the robots alone, before PCs, spare batteries and networking.

Is ROS 1 Noetic still supported?

No. ROS 1 support ended on May 31, 2025, and Noetic was the final ROS 1 distro. It still runs, but it gets no updates or security fixes.

Do I need a real robot to learn ROS?

No. Gazebo plus RViz2 covers topics, services, actions, TF, SLAM and navigation. Buy a robot once your code works in simulation and you need real sensor noise, latency and battery limits.

Where to start this week

Pull osrf/ros:jazzy-desktop-full (or the Lyrical equivalent), mount a workspace, and get the talker, RViz2 and an empty Gazebo world running. That gives you a working Tier 1 lab in an afternoon. Spend money on hardware only after your first package drives a simulated TurtleBot 4 around a map. By then you’ll know whether you need one Lite, a Standard or a whole room of them.

Credits and sources

By RWT Robotics Editorial Team

  • Facts were checked against the sources below before publication.
  • Published

Checking and sources

  • Researched from 8 sources, listed below.
  • Facts are checked against those sources; anything we cannot confirm is corrected, labelled or removed.
Sources (8)
  1. ROS 2 Release Schedule (Lyrical docs) (docs.ros.org)
  2. Lyrical Luth Supported Platforms (docs.ros.org)
  3. Release Lyrical Luth (docs.ros.org)
  4. ROS 2 Lyrical Luth testing kicks off on April 30th (discourse.openrobotics.org)
  5. ROS 2 Lyrical Luth and 11 years of Fast DDS as ROS 2 default middleware (discourse.openrobotics.org)
  6. Lyrical Luth Test and Tutorial Party Instructions (Docker images) (discourse.openrobotics.org)
  7. ROS 2 Lyrical vs Jazzy: Which LTS Should You Use in 2026? (openelab.io)
  8. ROS 2 end-of-life dates (endoflife.date)