Robotics · from zero to interview-ready
Learn robotics
in the right order.
Robotics is five disciplines braided together — math, mechanical, electronics, control, software. Most people get stuck not because it is hard, but because the order is wrong. This site flattens it: 74 topics, each stating what to learn, how deep each of the four career tracks needs it, and how to tell you have learned enough.
Career tracks
RM-01 The roadmap: four metro linesAll 74 stations in learning order, depths marked.
start here RM-02 Track 1 · Algorithm20 stations · ≈770 h. Perception, SLAM, motion control, VLA.
20 stations RM-03 Track 2 · System software18 stations · ≈685 h. Middleware, real-time, integration.
18 stations RM-04 Track 3 · Embedded & controls28 stations · ≈960 h. MCUs, motor drives, field buses.
28 stations RM-05 Track 4 · Mechanical10 stations · ≈405 h. Structure, transmission, lightweighting.
10 stationsLessons — where most people should start
L-01 Python & scientific computingPrototyping, data analysis, training scripts, visualization
30 h · Beginner L-02 Linear algebra & rigid-body rotationEvery coordinate transform a robot performs is matrix arithmetic
40 h · Intermediate L-03 Coordinate transforms & the tf treeThe shared spatial consensus of the whole robot system
15 h · Beginner L-04 Forward/inverse kinematics & JacobiansThe native language of a robot arm
40 h · Intermediate L-05 Classical control (PID / frequency domain / stability)80% of loops in industry still run PID
45 h · Intermediate L-06 Camera models & calibrationThe first gate of vision is not the model — it is calibration
20 h · Beginner L-07 ROS 2The industry's de facto standard
70 h · Intermediate L-08 Arduino & the maker ecosystemThe fastest way to build hands-on feel: plug in USB and it runs
15 h · Beginner L-09 Physics simulation (MuJoCo / Isaac)Trial and error in simulation costs three orders of magnitude less than crashing the
40 h · IntermediateInteractive labs — drag the physics
LAB PID Tuning LabDrag three gains and watch one control curve go from sluggish to oscillating and
8 min LAB The Jitter Bill of a 1kHz Control LoopThe same control loop on stock Linux vs PREEMPT_RT — how many microseconds apart,
6 min LAB What a Kalman Filter Actually DoesHow a noisy sensor reading becomes an estimate that is smooth without lagging behind
6 min LAB Gear Ratio to End-Effector ForceMotor gives 3 N·m; through a 100:1 reduction, how much can the end effector
4 minMath foundations — built from scratch
M The Matrix Is a Ledger for RotationYou already know that the derivative of f(x) = x² is 2x — and that sin(a+b) and cos(a+b)
≈ 25 min M Complex multiplication is rotation: the one-dimensional version of quaternionsYou learned complex multiplication in high school: (1+i)(1−i) = 2
About 20 min M Probability spaces: why «random» needs axiomsThree answers to the same problem means not that math failed, but that «random» was never defined
50 min M The Jacobian: Rotate a Joint One Degree — How Far Does the End Effector Move?The Jacobian is the conversion table for «each joint moves a little — how much does the end
About 45 minWhat is here, and how it is written
Every topic page answers three questions: what to learn, how deep each of the four career tracks requires it, and how to verify you have learned it. The same 74 topics also exist as 27 math derivation articles and 15 browser-run labs — no installation, no hardware, drag a slider and watch the physics.
Two editions share one editorial process and one set of sources: this English edition (robotguide.tech) and the Chinese edition (robotguide.dev). Data comes with its source; judgments are labeled as judgments. See how this site is built.