RobotGuide

For people who want to work in robotics

Read the map first.
Then decide what to learn.

Robotics is not one major. It is five disciplines braided together: math, mechanical, electronics, control, and software. Most people get stuck not because it is hard, but because the order is wrong — starting with a framework, then patching fundamentals, and ending up inside a black box.

RobotGuide flattens it: 74 topics, 4 career tracks, 8 job families, and interactive experiments you can drag in the browser.

74topics, each with acceptance criteria
4career tracks, pick one and go deep
~2685study hours from zero to interview-ready
90interactive figures you can drag

The problem this site solves

Entry-level robotics information is extremely fragmented. Half of it is university curricula, half is vendor keynotes. The part in between — in what order to learn, how deep is deep enough, what interviews actually ask — is barely covered anywhere in a systematic way.

That is the gap RobotGuide fills: for every topic, what to learn, how deep each of the four career tracks requires it, how to tell you have learned enough, and the mistakes people usually make on the way.

Four career tracks

The same 74 topics, cut four ways. Pick one track and go deep on it; shared stations are the common ground everyone needs.

Algorithm

Perception, SLAM, motion control, reinforcement learning, and VLA models. The highest salary ceiling, the steepest credential wall — most roles sit at the master's-degree level and above.

20 stations · ≈770 h

Stations →

System software

Middleware, real-time systems, architecture, integration. Fewer openings but chronically scarce: the layer that turns algorithms into something a real machine runs for a full day.

18 stations · ≈685 h

Stations →

Embedded & controls

MCUs, motor drives, FOC, sensors, field buses. Spans hardware and software; hard to replace the people who can hold the whole signal chain in their head.

28 stations · ≈960 h

Stations →

Mechanical

Body structure, transmission, lightweighting, thermal design. The lowest entry barrier, but the well-paid positions belong to people who understand the electronics and control around their parts.

10 stations · ≈405 h

Stations →

Seven domains

The same topics, organized by discipline. Every topic page states what to learn, how to verify you learned it, and the common mistakes.

Math foundations

5 topics

The shared base of every track. Time saved by skipping it is repaid tenfold later.

Robotics core

10 topics

The theory arms, quadrupeds, and humanoids all share. Skipping it for frameworks leaves you with black boxes.

Perception & intelligence

9 topics

Making robots see, understand, and decide. The most crowded, most expensive layer.

Systems & software

14 topics

Turning algorithms into something a real machine runs stably for a day.

Hardware & electrical

19 topics

Making things actually move. The closer to current and structure, the harder to replace.

Engineering practice

7 topics

Not taught in class, always asked in interviews. The difference between a course project and a product.

Tools & simulation

10 topics

Simulators, arm software stacks, optimization libraries, debugging and deployment tools.

What is where, right now

RobotGuide is published in two editions that share one editorial process and one set of sources: this English edition (robotguide.tech) and the Chinese edition (robotguide.dev). The complete library — all 74 topic pages, 90 interactive figures, 15 lab benches, six company architecture teardowns — currently lives on the Chinese edition. English pages are being added in batches, starting with this roadmap.

We only write what can be verified: data comes with its source, and judgments are labeled as judgments. See how this site is built.