Short, visual theory
One idea at a time: what an attack is, why it works, and what it would cost you. No walls of jargon.
Free interactive course · 10 lessons
Cyber Security
Cyber Security
for robotics & AI systems
Robots move through the world, talk over networks and trust their models. Every lesson puts you in the seat: trace an interception step by step, crack a cipher, watch an adversarial image flip a classifier, and defend the fleet.
- No install, no sign-up
- Simulated labs, safe & legal
- 24 interactive labs
Everything runs locally in your browser
simulated fleet · ready
10Lessons
4Parts
24Interactive labs
100%Free access
How it works
Read a little. Breach a lot — safely.
Security is learned by following an attack end to end, then defending against it. Each lesson alternates short explanations with simulations you control — no real targets, no risk.
Safe simulated labs
Tunnels, packet flows and terminals are fully simulated — deterministic, offline, and academically honest about what is real.
Step through attacks
Play, pause and scrub every lab: watch a packet take the attacker's route, a hash avalanche spread, defenses snap shut.
Check and practise
Explained quiz answers and a practical task with a hint and a worked solution close every lesson.
See a lab
Two exercises from the end of the course
A suspicious access log and a session-policy decision. You will analyse both types of artefact in the lessons that teach them.
10
One line gives the attacker away
Lesson 10 · detection & incident response
5
How long should a session last?
Lesson 5 · identity, passwords & access
Syllabus
Ten lessons, one continuous path
Four parts, each answering the question the previous part raises. Start at lesson 1 — every lesson assumes only the ones before it.
What you get
Offense to understand, defense to matter
When you finish lesson 10 you will not just know attack names. You will know how each one defeats a specific assumption — and which control removes that assumption.
Reason about threats
Assets, attack surfaces, STRIDE and risk trade-offs. You will look at any system — robot, cloud model, dashboard — and know what to question first.
Speak the language
Encryption vs hashing, authN vs authZ, TLS vs a VPN, IDS vs a firewall. Precision matters in security briefings — you will have it.
Secure the robot stack
Fleet networks, DDS/ROS 2, CAN bus, firmware signing and OTA updates — the full chain from Wi-Fi to the actuator.
Guard learning systems
Adversarial inputs, poisoned training sets and injected prompts — the failure modes unique to AI, plus the evaluation and monitoring that catch them.
“Every attack in this course defeats one specific assumption — that a packet is really from whom it claims, that a login is a person, that a photo is a scene. Security is the discipline of noticing assumptions, then removing the dangerous ones.”
Cyber Security for Robotics & AI · lesson 10
FAQ
Before you touch anything
The practical details.
Will I be hacking real systems?
No — and that is deliberate. Every lab is a simulation that runs entirely in your browser: fake networks, fake robots, deterministic outputs. You learn the exact mechanics attackers use, without touching a single real machine. Nothing here can be used against a third party, and that is a feature: it means the course can teach the whole attack chain honestly.
Do I need a robotics background?
No. The course starts from general security fundamentals and introduces every robotics concept it needs (DDS, CAN, OTA) from scratch. If you have read a Python script before you are fully equipped; if you also know ROS, the later lessons will feel unusually concrete.
What do I need to install?
Nothing. Terminals, packet flows, the cipher playground and the adversarial-image dial are all browser-native. Opening the site through a simple local web server is recommended so lesson scripts load cleanly, but any modern browser works.
Is my progress saved?
Yes, in this browser's localStorage — completed lessons and quiz results survive a reload but not a different device. No accounts, no servers, no tracking.
Where do I go after the course?
Lesson 10 ends with a concrete roadmap: labs to set up at home, certifications that are worth their cost, and how the skills map to security, robotics-ops and ML-engineering roles. Each lesson also links the primary sources worth reading next.
How long does the course take?
About 9–10 hours in total: 45–60 minutes per lesson, including the labs, quiz and practice task. One lesson per sitting works well; lessons 1–3 build the vocabulary the rest depends on.
Free interactive course
Your first threat model
Your first threat model
takes one lesson.
Then ten lessons take you from the CIA triad to defending an autonomous fleet — packet by packet, hash by hash. One idea, one lab, one decision at a time.
Free · No sign-up · Simulated & safe labs