Skip to content

Welcome to Fz2h's Robotics Course Website

This is a comprehensive engineering guide on Robot Operating System (ROS) + Visual Perception.

This course primarily explains how relatively novel algorithms are currently deployed on robots, aiming to provide assistance in course project design, graduation projects, research, and robotics introduction.

Regarding self-learning materials, although GPT/vibe-coding has greatly lowered the barrier to self-learning, for students with zero foundation, if they lack even the basic knowledge to ask questions and have insufficient ability to distinguish truth from falsehood in knowledge, this will lead to poor self-learning outcomes. Additionally, the lack of practical engineering goals makes it easy for people to give up learning and feel very lost during the learning process. Therefore, we design the course content with the goal of developing a relatively complete robotics project.

In principle, we don't reinvent the wheel and won't repeat tutorials that already exist. Instead, we'll provide links to relatively good tutorials. For example, there are currently many online ROS tutorials, so our tutorial won't cover too many ROS details, but will mention relatively key knowledge points and the overall engineering logic. Of course, this also depends on the response.

Video hands-on content will be updated on the bilibili channel:

Fz2h

Project Overview

The current project is to deploy object detection algorithms and frontier exploration algorithms on a single robot based on the ROS framework.

Course Modules

Robotics Basics

Mathematical foundations of robotics, covering forward/inverse kinematics, dynamics, coordinate transformations, and DH parameters.

ROS Robot Operating System

ROS communication mechanisms, robot environment configuration, and perception node development.

Perception Module

YOLO object detection, model training, and visual perception systems.

Simulation Environment

Simulation frameworks including Gazebo, Drake, MuJoCo, and PyBullet.

Robotic Manipulation (MIT 6.8210)

MIT Professor Russ Tedrake's manipulation course covering perception, planning, control, and reinforcement learning.

Planning Module

Task planning, temporal logic, reactive behaviors, and learning-enhanced planning.

Reinforcement Learning

RL fundamentals, value functions, policy gradients, and model predictive control.

Deployment

Deploying from simulation to real robots, hardware integration, system architecture, and safety considerations.

Course Roadmap

Preliminary

Parallel Learning Path

Kroki

Course Update Log

  • 2026-04-26: Updated Reinforcement Learning module
  • 2026-04-26: Updated Robotic Manipulation module (MIT 6.8210)
  • 2026-04-26: Added Planning module (task planning, temporal logic)
  • 2026-04-26: Updated Simulation module (Gazebo, Drake, MuJoCo, PyBullet)
  • 2026-04-26: Updated Deployment module (Sim-to-Real, hardware integration)
  • 2025-12-20: Built visual perception node
  • 2025-12-18: Linux quick start
  • 2025-12-17: Environment setup: Python
  • 2025-12-16: Initialized ROS tutorial