Robotics
Physical AI: Teach a Robot Arm with LeRobot
Teach a robot arm by showing it, using LeRobot, imitation learning and simulation, and learn the safety limits of learned control.
Physical AI: Teach a Robot Arm with LeRobot is a free, self-paced online course from EDWartens for robotics, mechatronics and automation engineers and engineering students worldwide who want to teach robot arms from demonstrations, with or without hardware. It has 13 modules and takes about 17.5 hours, with written notes, a practice task per module, a practical project and a final assessment of 15 questions (pass mark 60%). Learning costs nothing and needs only a free account. The optional EDWartens Certificate of Completion, verifiable by code, is a one-off US$28.99 for you; it is not a vendor credential. Updated 1 November 2026.
- Free
- Intermediate
- 17.5 hours
- Self-paced
- Certificate optional
- Lab practice in your browser
No card, no trial. The course is free for good. Certificate US$28.99, optional.
- Who it is for
- Robotics, mechatronics and automation engineers and engineering students worldwide who want to teach robot arms from demonstrations, with or without hardware
- Jobs
- Roles that use this skill: robot programmer, automation engineer, cell integrator and maintenance technician. No course guarantees a job.
- Tools
- LeRobot (open source, Apache 2.0, version 0.6 line in October 2026) with Python 3.12 and PyTorch, Google Colab for a free GPU, the PushT and ALOHA simulation environments, and the Hugging Face Hub. Isaac Sim is free but optional and needs an RTX GPU. Hardware: none required: every module has a simulation-only path that runs on Colab. Optional: an SO-101 leader and follower arm kit with two USB cameras, and an NVIDIA GPU with 8 GB or more for local training. Isaac Sim needs a recent RTX GPU with about 16 GB of memory.
- Updated


After this course you can
By the end you will be able to:
- Explain how a learned policy differs from a taught-point program
- Load and inspect LeRobot datasets from the Hugging Face Hub
- Train an ACT policy on a simulated task
- Teleoperate an SO-101 arm and record demonstrations
- Evaluate a policy with an honest success rate
Who it is for, and what you need first
Who it is for
Robotics, mechatronics and automation engineers and engineering students worldwide who want to teach robot arms from demonstrations, with or without hardware.
Level: Intermediate. Intermediate. For robotics, mechatronics and automation engineers and students who can run a Python notebook. No machine learning background is needed beyond what the course explains; Python for AI and Engineering Data is a good course to take first if Python is new.
What you need
Software: LeRobot (open source, Apache 2.0, version 0.6 line in October 2026) with Python 3.12 and PyTorch, Google Colab for a free GPU, the PushT and ALOHA simulation environments, and the Hugging Face Hub. Isaac Sim is free but optional and needs an RTX GPU..
Hardware: None required: every module has a simulation-only path that runs on Colab. Optional: an SO-101 leader and follower arm kit with two USB cameras, and an NVIDIA GPU with 8 GB or more for local training. Isaac Sim needs a recent RTX GPU with about 16 GB of memory.. What to download, and how.
How this course works
Watch the lessons
Short video lessons, one module at a time, at your own pace.
Read the notes
Open the Reading tab for the written notes and diagrams of the module.
Practise in the 3D labOptional
Try the module in the free 3D lab, in your browser.
Check yourself
Answer the questions to see what stuck before you move on.
Final assessment
Pass the final once you have been through the modules.
Certificate, if you want oneOptional
Optional, with a page anyone can use to verify it.
What you learn
13 modules in order, at your own pace. 17.5 hours of video lessons, with written notes and a practice task in each module.
- 1From programmed robots to learned skills: what physical AI is, with lab practice: 31m
- 2The LeRobot stack and datasets on the Hugging Face Hub42m
- 3Imitation learning: behaviour cloning and why it drifts59m
- 4Simulation first: gym environments and PushT on Colab28m
- 5ACT: action chunking with transformers, explained1h 9m
- 6Diffusion policy and choosing a policy1h 14m
- 7SO-101 hardware: motors, assembly and calibration (optional)27m
- 8Teleoperation and recording demonstrations, with lab practice: 35m
Show the other 5 modules
- 9Training on real data, deploying and evaluating1h 54m
- 10Vision-language-action models and SmolVLA52m
- 11Isaac Sim and sim-to-real: an overview (optional, needs an RTX GPU)46m
- 12Safety: cobot standards, guarding and where learned control is not allowed, with lab practice: 52m
- 13Final Assessment2m
Software you need
What to download, where from, what it costs and how to install it. Every link goes to the maker's own site, never a mirror.
LeRobot installs with pip (pip install lerobot, plus the pusht, aloha or feetech extras) and needs Python 3.12 or later. The simulation path runs entirely on Colab. Isaac Sim is optional and needs an RTX GPU.
Show what to download (4 tools)
Required
- 01Free, GPU time not guaranteed
Google Colab
Google, in the browser
- Runs on
- Any modern web browser
- Account
- A free Google account
Colab is free to use. In the free version GPUs and TPUs are heavily restricted and not guaranteed, sessions can run for at most 12 hours, and idle sessions are stopped. Paid plans give more reliable access.
Steps
- 1.Open colab.research.google.com and sign in with your Google account.
- 2.Click New notebook (or open the notebook your course links to).
- 3.For a GPU, choose Runtime, then Change runtime type, then pick a GPU if one is offered.
- 4.Type code in a cell and press Shift and Enter to run it.
- Save a copy to your Google Drive so your changes are kept.
- Files on the runtime are deleted when the session ends. Save outputs to Drive.
Open Google Colabcolab.research.google.com
Optional
Useful, not needed to finish the course.
- 02Free
Python
Python Software Foundation
- Runs on
- Windows (not Windows 7 or earlier), macOS and Linux. Windows builds for x64, 32-bit and Arm64.
- Account
- None needed
Python is free, open source software. You can use it for learning and for commercial work at no cost.
Steps
- 1.Open python.org/downloads and click the download button for your system.
- 2.On Windows, run the Python install manager (or the classic 64-bit installer) you downloaded.
- 3.If you use the classic installer, tick "Add python.exe to PATH" on the first screen, then click Install Now.
- 4.Open a new Command Prompt or Terminal and run: python --version (on Windows you can also run: py --version).
- 5.Install packages with pip, for example: python -m pip install requests
- pip comes with Python. Run it as python -m pip so it always matches the Python you are using.
- Python.org now recommends the Python install manager on Windows. If it offers to add its folder to PATH, say yes so the python command works everywhere.
- Make a virtual environment for each project: python -m venv .venv
Official download pagepython.orgAlternatives
- Anaconda Distribution: Python with 600+ data science packages and Jupyter already included.
- 03Free
Git
Git project (git-scm.com)
- Runs on
- Windows (x64 and ARM64), macOS, Linux
- Account
- None needed
Git is free and open source software. Use it for any purpose at no cost.
Steps
- 1.Open git-scm.com/downloads and choose your system.
- 2.On Windows, download the x64 Setup (or ARM64 Setup) and run it. The default choices are fine for most learners.
- 3.Open a new terminal and run: git --version
- 4.Set your name: git config --global user.name "Your Name"
- 5.Set your email: git config --global user.email "you@example.com"
- On Windows you can also install with: winget install --id Git.Git -e --source winget
- A portable version (no installer) is offered for PCs where you cannot install software.
Official download pagegit-scm.com - 04Free
Visual Studio Code
Microsoft
- Runs on
- Windows 64-bit (supported Windows client versions), macOS (latest and two previous releases), Linux (Ubuntu 20.04, Debian 10, RHEL 8, Fedora 36 or later)
- Account
- None needed
- Size
- Less than 200 MB download, under 500 MB installed
Free to download and use. Extensions from the Marketplace each have their own licence.
Steps
- 1.Open code.visualstudio.com/download.
- 2.Pick the installer for your system (on Windows, the User Installer is fine).
- 3.Run the installer. On Windows, tick "Add to PATH" if it is offered.
- 4.Open VS Code and install the extensions your course uses, for example the Python extension.
- Hardware needs are small: a 1.6 GHz processor and 1 GB of RAM.
- On Windows, the "Open with Code" options in the installer let you open folders from File Explorer.
Official download pagecode.visualstudio.com
Checked against each maker's own page on 27 September 2026. Trial lengths and editions change; the maker's page is the final word.
Practise in the 3D lab
Teach an arm the classical way, with taught points, and prove its guards stop it, so you can judge what a learned policy adds and what it must never replace.
- JogRobotics lab, about 20 minOpen in the lab (opens in a new tab): Jog
- Teach and runRobotics lab, about 30 minOpen in the lab (opens in a new tab): Teach and run
- Safety validationRobotics lab, about 25 minOpen in the lab (opens in a new tab): Safety validation
Optional. Skip it whenever you like: it never blocks your progress or your certificate.
Course project, about 16 hours
Train and honestly evaluate a learned policy on PushT, with an optional SO-101 pick-and-place
Evaluate a pretrained diffusion policy on the PushT simulator as a baseline. Then train your own diffusion and ACT policies on the same data within a fixed GPU budget. Evaluate each on 100 episodes with a written protocol. Report every success rate with a 95 percent interval. Write a recommendation that says only what the numbers support. Optionally repeat the evaluation on an SO-101 pick-and-place.
- 5documents in the pack
- 11steps to follow
- 8things to hand in
- 16 hof work, at your pace
- PlanStudy plan: learned policies in simulation before a bench trial
- ProcedureEvaluation procedure for learned policies on PushT
- Test reportPushT evaluation of four policies
- Risk registerBench risk register for the optional SO-101 trial
- ReportRecommendation on learned policies for the kitting enquiry
The project is optional practice. The certificate needs only the modules and the final assessment.
After the course, your next step
Show what you can do, then keep going in the same field.
A certificate you can verify
Finish the modules and pass the final assessment (15 questions, pass mark 60%) and you can add a certificate with a public verification page. One-off, US$28.99 for you. It is not a vendor credential.
Want a free certificate? Do the matching Free internship: 4 weeks, nothing to buy, with a free certificate of completion.

PLC Programming
Eight short weeks of PLC ladder logic on free simulators: notes, diagrams, a few videos and auto-tested tasks. Finish weeks 1 to 4 and the final, then claim a free Internship Completion Certificate with your real dates and study hours, and a QR code anyone can scan to verify it. Training, not employment.
See the internshipMore free courses
Roles that use this skill
- Robot programmer
- Automation engineer
- Cell integrator
- Maintenance technician
These are the kinds of roles where the skill is used. We make no promise of a job or an interview.
Plain notes
- Free: the whole course, the notes, the practice tasks, the final assessment and support from practising engineers.
- Paid, and optional: the course certificate, and the Engineer-verified internship (the Free internship costs nothing).
- No job promise: the course teaches a skill. It does not place you in a job.
- Lessons by: Hugging Face, NVIDIA, IBM Technology, Google for Developers, RAIL, e-Yantra, Welch Labs, Trelis Research, Articulated Robotics, exida, Ilia, Phospho AI, Sanjiban Choudhury, Aleksandar Haber PhD, Foundation Models For Robotics, Trossen Robotics, Robraintics, Trushant Adeshara, Pius Lim, Shane Reetz, Lightwheel, Learning Automation. Independent creators; the lessons stream from their own channels. EDWartens wrote the study plan, notes, practice tasks and assessments.
Hugging Face · NVIDIA · IBM Technology · Google for Developers · RAIL · e-Yantra · Welch Labs · Trelis Research · Articulated Robotics · exida · Ilia · Phospho AI · Sanjiban Choudhury · Aleksandar Haber PhD · Foundation Models For Robotics · Trossen Robotics · Robraintics · Trushant Adeshara · Pius Lim · Shane Reetz · Lightwheel · Learning Automation. If you are one of these creators and want a lesson removed or credited differently, write to info@edwartens.com.
Practise what you learn in the free 3D lab
Technical work is judged on what you can do: wiring, commissioning, finding a fault. Doing a task yourself teaches more than watching it a second time, and a lab lets you try it safely before you meet real equipment. It does not replace working on real hardware.
- Practise safely, so a mistake costs nothing.
- The lab checks your work like a site test, so you know when it is right.
- Free, in your browser, with nothing to install.
Why learn with EDWartens
What you get with this course, and what comes after it.
- Free from the first lesson to the final assessment.
- Practice in a free 3D lab in your browser, with nothing to install.
- Video lessons with written notes and a project.
- An optional certificate that anyone can verify online.
- Free internships to keep practising afterwards.
Common questions
Who is this physical AI and LeRobot course for?
It is for robotics, mechatronics and automation engineers and engineering students. They want to move from programming robots point by point to teaching them from demonstrations. Robot programmers from FANUC, ABB, KUKA or Universal Robots backgrounds, and ROS 2 users, will find it a natural next step.
Do I need a robot arm to take the course?
No. Every module has a simulation-only path. The PushT and ALOHA simulators run on a free Google Colab GPU. Real-robot datasets, including SO-101 ones, are on the Hugging Face Hub to train on. The SO-101 modules are optional. They show what to do if you buy or build an arm later.
What do I need to know first?
You should be able to run a Python notebook and follow a command line. The course explains the machine learning it uses. If Python is new, take Python for AI and Engineering Data first; for classical robot software, ROS 2 Robot Programming with Python complements this course.
Which software does it use?
LeRobot, Hugging Face's open-source robot learning library (version 0.6 line in October 2026, Python 3.12 or later), with PyTorch, Google Colab and the Hugging Face Hub. Isaac Sim appears in one optional module and needs an RTX GPU. LeRobot changes quickly, so the notes give the current commands and tell you to check the release notes.
Is the course free to learn?
Yes. Every module, the notes, the worked problems, the project and the final assessment are free, and all the software used is free. The certificate is optional.
How long does the Physical AI: Teach a Robot Arm with LeRobot course take?
About 17 to 18 hours of video, notes and practice at your own pace, of which about 10.5 hours is video. The optional PushT project takes about 16 hours more, most of it waiting for training runs.
Is a learned policy safe to use on a factory robot?
Only for task decisions, inside the robot's rated safety configuration and a proper risk assessment. A learned policy is never allowed to be a safety function. The safety module covers ISO 10218-1 and -2 (2025 editions), why protective stops and speed limits must be rated hardware, and how to run a first test safely.
What certificate does the Physical AI: Teach a Robot Arm with LeRobot course give?
An EDWartens certificate of completion. It is issued when you finish the modules and pass the 15-question final at 60 percent. It has a number anyone can verify on our site. It is not a credential from Hugging Face, NVIDIA or any robot maker. None of them is affiliated with EDWartens.
Do I have to use the lab?
No. The lab is optional practice. You can skip it whenever you like, and it never blocks your progress, the final assessment or your certificate.
Is the lab free?
Yes. Every lab level for this course opens in your browser, free, with nothing to install.
What topics does the Physical AI: Teach a Robot Arm with LeRobot course cover, module by module?
The Physical AI: Teach a Robot Arm with LeRobot course has 13 modules. It opens with "From programmed robots to learned skills: what physical AI is", goes through "The LeRobot stack and datasets on the Hugging Face Hub", "ACT: action chunking with transformers, explained", "Training on real data, deploying and evaluating" and "Safety: cobot standards, guarding and where learned control is not allowed" and ends with "Final Assessment". The full list with minutes is under "What you learn" on this page.
What can I do after the Physical AI: Teach a Robot Arm with LeRobot course?
After the Physical AI: Teach a Robot Arm with LeRobot course: Explain how a learned policy differs from a taught-point robot program, and where each belongs; Load LeRobot datasets from the Hugging Face Hub and read their observations and actions; Explain behaviour cloning, compounding error and why action chunking and diffusion help. These are the practical outcomes the course is written around, and the final assessment of 15 questions checks them.
What software and hardware do I need for the Physical AI: Teach a Robot Arm with LeRobot course, and does it cost anything?
For the Physical AI: Teach a Robot Arm with LeRobot course, software: LeRobot (open source, Apache 2.0, version 0.6 line in October 2026) with Python 3.12 and PyTorch, Google Colab for a free GPU, the PushT and ALOHA simulation environments, and the Hugging Face Hub. Isaac Sim is free but optional and needs an RTX GPU. Hardware: none required: every module has a simulation-only path that runs on Colab. Optional: an SO-101 leader and follower arm kit with two USB cameras, and an NVIDIA GPU with 8 GB or more for local training. Isaac Sim needs a recent RTX GPU with about 16 GB of memory. The course itself is free.
Is the Physical AI: Teach a Robot Arm with LeRobot course suitable for me, and what level is it?
Physical AI: Teach a Robot Arm with LeRobot is rated intermediate. Intermediate. For robotics, mechatronics and automation engineers and students who can run a Python notebook. No machine learning background is needed beyond what the course explains; Python for AI and Engineering Data is a good course to take first if Python is new. It is written for robotics, mechatronics and automation engineers and engineering students worldwide who want to teach robot arms from demonstrations, with or without hardware, and takes about 17.5 hours at your own pace.
Which jobs use the skills from the Physical AI: Teach a Robot Arm with LeRobot course?
Skills from the Physical AI: Teach a Robot Arm with LeRobot course (for example: explain how a learned policy differs from a taught-point robot program, and where each belongs) are used in roles such as robot programmer, automation engineer, cell integrator and maintenance technician, and in robotics engineer work. The course gives you practical evidence of the skill; it does not promise a job, and employers decide on experience and interview.
Start free today
Physical AI: Teach a Robot Arm with LeRobot, at your own pace. Add the certificate only if you want one.