Robotics + AI, from first circuit to autonomous system

Design. Build. Program. Automate.

A structured path through robotics engineering, electronics, computer vision, and embedded AI — with real projects, simulations, and an assistant that helps you debug, not just copy.

View the roadmap → Browse projects

8Learning domains

3Skill tiers

25Book summaries

28Tools covered

🐍 Python · ROS 2

👁️ OpenCV live feed

⚙️ ESP32 flashed

Who it's for

Wherever you're starting from

🎓

Curious Beginner

New to robotics. Starts with math, physics, and a first Arduino build.

🛠️

Hobbyist Maker

Builds for fun. Wants project templates and community challenges.

📐

CS / EE Student

Needs ROS 2, embedded systems, and CV projects for a placement portfolio.

🚀

Career Switcher

Moving into automation or AI engineering — follows the full certification path.

Learning domains

Eight paths into robotics and AI engineering

Each hub pairs core theory with the tools professionals actually use.

🤖

Robotics Fundamentals

Robot types, kinematics, mechanics, sensors & actuators, motion systems

KinematicsMotion systemsSensors

Tools — Arduino IDE, PlatformIO, Fusion 360, SolidWorks, Tinkercad

🧠

AI for Robotics

Machine learning, computer vision, object detection, speech, decision-making

MLObject detectionSpeech

Tools — Python, TensorFlow, PyTorch, OpenCV, MediaPipe, YOLO

⚙️

Electronics & Embedded Systems

Arduino, ESP32, Raspberry Pi, microcontrollers, motor drivers, PCB design

ESP32PCB designMotor drivers

Tools — Arduino IDE, KiCad, STM32CubeIDE, PlatformIO

👁️

Computer Vision & Navigation

Camera integration, face detection, SLAM, path planning, sensor fusion

SLAMPath planningSensor fusion

Tools — OpenCV, ROS 2, Gazebo, RViz

🦾

Robot Arm & Industrial Robotics

Robotic arms, inverse kinematics, PLC basics, factory automation

Inverse kinematicsPLCPick-and-place

Tools — ROS 2, MoveIt, Siemens TIA Portal, CoppeliaSim

🏠

IoT & Smart Robotics

Smart home robots, IoT devices, cloud connectivity, MQTT, edge AI

MQTTEdge AICloud

Tools — ESP32, MQTT, Node-RED, Firebase, Blynk

💻

Programming & Simulation

Python, C++, ROS 2, Linux, simulation environments, testing

ROS 2LinuxTesting

Tools — VS Code, Ubuntu, ROS 2, Gazebo, Webots

🚀

Projects & Competitions

Line followers, obstacle avoidance, AI vision robots, delivery robots

ChallengesDrone conceptsRobot dogs

Tools — GitHub, Docker, ROS 2, Fusion 360, Gazebo

Course structure

One curriculum, traced like a circuit

Each tier builds on the last — the same way a signal moves through a board, one verified stage at a time.

Mathematics & Physics

Vectors, linear algebra, kinematics groundwork.

Basic Electronics

Circuits, voltage, current, breadboarding.

Arduino Programming

Digital I/O, PWM, your first sketch.

Python

Syntax, control flow, scripting for hardware.

Mechanical Design

Chassis basics in Tinkercad and Fusion 360.

Introduction to AI

What ML is and where it fits in robotics.

Sensors & Actuators

IMUs, encoders, servos, and motor control.

Embedded Systems

ESP32 and Raspberry Pi as robot controllers.

Computer Vision

OpenCV pipelines, detection basics.

Robot Navigation

Path planning and obstacle handling.

CAD Modeling

Full chassis and mount design.

ROS 2 Basics

Nodes, topics, and your first ROS 2 package.

Autonomous Robots

SLAM, sensor fusion, closed-loop autonomy.

Multi-Robot Systems

Coordination and communication at scale.

Industrial Automation

PLCs, pick-and-place, factory workflows.

AI Integration

Deploying trained models onto real robots.

Edge AI · Cloud Robotics

On-device inference and cloud-connected fleets.

Digital Twins · System Design

Simulating and architecting full robotic systems.

Checkpoint 01

🥉

Robotics Foundations

Awarded after completing the Beginner tier and one starter project.

Checkpoint 02

🥈

Embedded Systems Engineer

Awarded after Sensors, Embedded Systems, and ROS 2 Basics.

Checkpoint 03

🥇

Computer Vision Engineer

Awarded after the CV module and an AI Vision Robot build.

Checkpoint 04

🏆

Autonomous Systems Engineer

Awarded after completing the full Advanced tier and capstone.

Learning hub

Everything around the curriculum

Roadmaps, references, and tooling in one place.

Beginner-to-advanced roadmaps

A visual map of every tier, so you always know the next node.

Top 25 book summaries

Structured breakdowns of the field's essential references.

Official tool setup guides

Step-by-step installs for every tool used across the curriculum.

Hands-on labs & simulations

Practice in Gazebo, Webots, and CoppeliaSim before touching hardware.

CAD design tutorials

Model your own robot chassis in Fusion 360, SolidWorks, or Tinkercad.

AI coding assistant

Get unstuck on Arduino, Python, or ROS 2 code without losing the lesson.

Project-based learning

Every module ends in something you actually built and can show.

Progress dashboard

Track modules, skills, and portfolio pieces in one view.

Certification roadmaps

See exactly which skill checkpoints map to industry credentials.

Tool setup guides

Every tool, one setup guide away

OS-specific install steps for the software and platforms used across all eight domains.

🔧

Arduino IDE

🔌

PlatformIO

🧊

Fusion 360

📐

SolidWorks

🧱

Tinkercad

🐍

Python

🔶

TensorFlow

🔥

PyTorch

👁️

OpenCV

✋

MediaPipe

🎯

YOLO

🖥️

KiCad

💠

STM32CubeIDE

🦾

ROS 2

🌍

Gazebo

📊

RViz

🤏

MoveIt

🏭

Siemens TIA Portal

🎛️

CoppeliaSim

📶

MQTT

🔴

Node-RED

🔥

Firebase

🔵

Blynk

📝

VS Code

🐧

Ubuntu

🕹️

Webots

🐙

GitHub

🐳

Docker

Projects & competitions

Build something that moves

Starter builds with a bill of materials, wiring guide, and starter repo.

🛞

Beginner

Line Follower Robot

IR sensors and a simple control loop — your first closed-loop robot.

ArduinoIR sensors

BOM: 6 partsStarter repo →

View build guide

Bill of materialsArduino Uno/Nano · 5-channel IR sensor array · 2× DC gear motors · L298N motor driver · 2WD chassis kit · battery pack

Bolt the two DC motors and caster wheel onto the chassis, then mount the IR sensor array facing down at the front edge.
Wire the L298N driver's motor outputs to the two DC motors, and its input pins to 4 digital pins on the Arduino.
Connect the IR array's output pins to 5 analog/digital pins and power it from the 5V rail.
Flash the starter sketch: read all 5 sensors, compute an error value from which sensors see the line.
Apply proportional correction — turn power toward whichever side reads "off line" — and tune the gain until it tracks smoothly.
Test on a taped track with a straight section first, then curves, then a 90° turn and an intersection.

📡

Beginner

Obstacle Avoidance Robot

Ultrasonic ranging and reactive navigation logic on ESP32.

ESP32Ultrasonic

BOM: 7 partsStarter repo →

View build guide

Bill of materialsESP32 dev board · HC-SR04 ultrasonic sensor · 2× DC motors · L298N motor driver · chassis kit · servo pan mount (optional) · battery pack

Assemble the chassis and mount the ultrasonic sensor facing forward, optionally on a servo for scanning.
Wire the HC-SR04's Trig and Echo pins to two ESP32 GPIOs, and the motor driver inputs to four more.
Flash firmware that pings the sensor every 100ms and converts the echo time into a distance in cm.
Add reactive logic: drive forward while distance is above a safe threshold, then stop, reverse briefly, and turn when it isn't.
Calibrate the threshold distance and turn duration to match your motor speed and chassis width.
Test in an open space with random obstacles, then a narrow corridor to check turning clearance.

👁️

Intermediate

AI Vision Robot

Real-time object detection driving motor commands via OpenCV + YOLO.

OpenCVYOLO

BOM: 9 partsStarter repo →

View build guide

Bill of materialsRaspberry Pi 4 (or Jetson Nano) · Pi Camera module · 2× DC motors · motor driver · chassis kit · battery pack · microSD card · cooling fan

Flash Raspberry Pi OS, enable the camera interface, and confirm a live feed with a short test script.
Install OpenCV and a lightweight YOLO variant sized for on-device inference.
Run detection on live frames and draw bounding boxes to confirm objects are recognized correctly.
Pick one target class (e.g. a ball or person) and compute how far its bounding box center is from the frame center.
Map that offset to motor commands — turn toward the object, drive forward as it grows closer in frame.
Set the script to run on boot and test end-to-end tracking across different lighting conditions.

📦

Intermediate

Delivery Robot

Path planning and SLAM concepts for a small indoor delivery platform.

ROS 2SLAM

BOM: 12 partsStarter repo →

View build guide

Bill of materialsRaspberry Pi 4 · 2D LIDAR (or depth camera) · IMU · 4× DC motors + encoders · motor controller · chassis kit · battery pack · Wi-Fi module

Install Ubuntu and ROS 2 on the Pi, then set up a workspace with navigation and SLAM packages.
Wire the LIDAR, IMU, and motor encoders, and publish their readings as ROS 2 topics.
Teleoperate the robot around the delivery area once to build a map using a SLAM package.
Save the map and configure the navigation stack's costmaps and path planner around it.
Set named delivery waypoints (e.g. "reception", "desk 3") and test autonomous navigation between them.
Layer in dynamic obstacle handling so the robot re-routes around anything not on the saved map.

🚁

Advanced

Drone Concepts

Flight dynamics and control theory, explored first in simulation.

GazeboControl theory

Sim-firstStarter repo →

View build guide

Simulation stackGazebo · a drone SITL package (e.g. PX4 or ArduPilot) · Python or C++ control scripts · optional companion computer for later hardware

Install Gazebo and a drone simulation package, then launch the default quadcopter model to confirm it spawns and responds to basic commands.
Study the four control axes — throttle, pitch, roll, yaw — and how each maps to motor speed differences.
Implement a basic PID controller for altitude hold and tune it until the drone hovers at a stable height.
Script a takeoff → hover → land sequence and verify it completes smoothly without oscillation.
Add simple waypoint navigation, sending the drone between 3–4 fixed points in the simulated world.
Once stable in simulation, review sim-to-real considerations (weight, ESC calibration, fail-safes) before touching real hardware.

🐕

Advanced

Robot Dog Concepts

Legged locomotion and gait planning for quadruped platforms.

CoppeliaSimGait planning

Sim-firstStarter repo →

View build guide

Simulation stackCoppeliaSim · a quadruped robot model (12-DOF legs) · Python or Lua control scripts

Install CoppeliaSim and load a quadruped model, confirming each leg's joints respond to direct angle commands.
Study common gaits — trot, walk, bound — and pick trot as the simplest starting pattern (diagonal leg pairs move together).
Implement inverse kinematics for one leg: given a foot position, solve for the hip, knee, and ankle joint angles.
Apply the same IK function to all four legs and script a basic trot cycle, checking the body stays level.
Tune step height, stride length, and cycle speed until the walk is stable at a moderate pace.
Test turning by varying stride length side-to-side, and note what would need to change for real hardware (motor torque, weight).

Zero to Advanced Masterclass

Step-by-step: Build one robot, evolve it infinitely

Follow this track to build a chassis from scratch, program its first movements in C++, add an AI vision brain in Python, and finally give it autonomous SLAM navigation via ROS 2.

Hardware & Logic

We begin by assembling the basic 2WD chassis and wiring the L298N motor driver to an Arduino. This establishes the low-level motor control loop.

Assemble chassis and mount motors, Arduino, and L298N.
Wire L298N IN1, IN2, IN3, IN4 to Arduino digital pins 8, 9, 10, 11.
Wire ultrasonic sensor Trig to pin 12, Echo to pin 13.
Download the CH340 serial driver (if using a clone board) and Arduino IDE.
Compile and upload the obstacle avoidance sketch.

⬇ Arduino IDE ⬇ CH340 Driver

// Stage 1: Basic Reactive Obstacle Avoidance (C++) const int trigPin = 12; const int echoPin = 13; const int IN1 = 8, IN2 = 9, IN3 = 10, IN4 = 11; void setup() { pinMode(trigPin, OUTPUT); pinMode(echoPin, INPUT); pinMode(IN1, OUTPUT); pinMode(IN2, OUTPUT); pinMode(IN3, OUTPUT); pinMode(IN4, OUTPUT); } void loop() { long duration, cm; digitalWrite(trigPin, LOW); delayMicroseconds(2); digitalWrite(trigPin, HIGH); delayMicroseconds(10); digitalWrite(trigPin, LOW); duration = pulseIn(echoPin, HIGH); cm = duration / 29 / 2; if (cm < 20) { // Obstacle detected: turn right digitalWrite(IN1, HIGH); digitalWrite(IN2, LOW); digitalWrite(IN3, LOW); digitalWrite(IN4, HIGH); } else { // Clear path: forward digitalWrite(IN1, HIGH); digitalWrite(IN2, LOW); digitalWrite(IN3, HIGH); digitalWrite(IN4, LOW); } }

The AI Brain Upgrade

We add a Raspberry Pi and a camera. The Arduino becomes a slave controller, accepting Serial commands from a Python script running YOLOv8 object detection on the Pi.

Flash Raspberry Pi OS using Pi Imager.
Connect Pi to Arduino via USB.
Install Python, OpenCV, and the Ultralytics YOLO library.
Flash the Arduino with a "Serial Listener" sketch.
Run the Python vision script to track a target.

⬇ Pi Imager ⬇ Python 3 ⬇ YOLO Weights

# Stage 2: AI Vision Tracking (Python + OpenCV) import cv2 import serial from ultralytics import YOLO # Connect to Arduino ser = serial.Serial("/dev/ttyUSB0", 9600) model = YOLO("yolov8n.pt") # Load nano model cap = cv2.VideoCapture(0) while True: ret, frame = cap.read() results = model(frame, classes=[0]) # Track class 0 (Person) if len(results[0].boxes) > 0: box = results[0].boxes[0].xyxy[0] cx = int((box[0] + box[2]) / 2) # Frame center is ~320 (assuming 640x480) if cx < 250: ser.write(b'L') # Turn Left elif cx > 390: ser.write(b'R') # Turn Right else: ser.write(b'F') # Forward else: ser.write(b'S') # Stop if nothing detected cv2.imshow("Vision", results[0].plot()) if cv2.waitKey(1) & 0xFF == ord('q'): break

ROS 2 Autonomy & SLAM

We replace the simple Python script with a full ROS 2 framework. A 2D LIDAR is added to scan the room, allowing the robot to map its environment and navigate autonomously.

Install Ubuntu 22.04 on the Pi, followed by ROS 2 Humble.
Mount an RPLIDAR A1 to the top tier of the chassis.
Setup micro-ROS on the Arduino to publish hardware encoder ticks.
Launch slam_toolbox to generate a map from LIDAR scans.
Use Nav2 to send goal poses to the robot dynamically.

⬇ Ubuntu 22.04 ⬇ ROS 2 Humble ⬇ Nav2 Package

# Stage 3: ROS 2 Launch File (Python) from launch import LaunchDescription from launch_ros.actions import Node def generate_launch_description(): return LaunchDescription([ # Start LIDAR Node Node( package='rplidar_ros', executable='rplidar_composition', name='rplidar_node', parameters=[{ 'serial_port': '/dev/ttyUSB1', 'frame_id': 'laser_frame', 'angle_compensate': True }] ), # Start SLAM Toolbox Node( package='slam_toolbox', executable='async_slam_toolbox_node', name='slam_toolbox', parameters=[ './config/mapper_params_online.yaml' ] ) ])

My ESP32 keeps disconnecting from MQTT every few minutes. Here's my code.

Your keep-alive interval is longer than the broker's timeout. Try lowering it to 15s and check the Wi-Fi power-save mode — that's the usual cause on ESP32.

Embedded Systems · IoT & Smart Robotics

Why does my SLAM map drift after a few loops?

Likely accumulated odometry error. Add loop closure or fuse in IMU data — want a short sensor-fusion example first?

Computer Vision & Navigation

AI coding assistant

Debugging help, not copy-paste answers

Context-aware

Understands which lesson, tool, or project you're in.

Explains, then suggests

Walks through the likely cause before giving a fix.

Covers the whole stack

Universal Build Blueprint

Sourcing & CAD

Gather components (motors, controllers, sensors) and 3D print or assemble the chassis framework.

Mechanical Assembly

Mount actuators to the frame. Ensure structural stability and test passive joint movements.

Electronics & Wiring

Connect motor drivers, sensors, and power distribution to your microcontroller (e.g., Arduino/ESP32).

Low-Level Logic (C++)

Write firmware to handle sensor interrupts, PID motor control, and basic reactive behaviors.

High-Level AI (Python/ROS)

Integrate a companion computer (Raspberry Pi/Jetson) for computer vision, mapping, and deep learning.

Engineering Stack

Tools & Setup Hub

The software and hardware toolkit required to build modern autonomous systems.

🔌

Arduino IDE

Setup: Download installer, run, install CH340 drivers if using clone boards.

The standard editor for writing C++ firmware for Arduino and ESP32 microcontrollers. Handles low-level motor control.

⬇ Download IDE

🐍

Python 3

Setup: Check "Add to PATH" during Windows install. Use pip install opencv-python.

Used for high-level scripting, computer vision (OpenCV), and interfacing AI models with serial connections.

⬇ Download Python

🐢

ROS 2 Humble

Setup: Requires Ubuntu 22.04. Run setup scripts from the official ROS 2 documentation.

The Robot Operating System. The industry standard framework for SLAM, navigation, and distributed node communication.

⬇ Setup ROS 2

🤖

Gazebo Simulator

Setup: Usually bundled with ROS 2 desktop-full installation.

A robust physics simulator. Test your robotic algorithms in a virtual environment before risking real hardware.

⬇ Install Gazebo

📐

Fusion 360 / FreeCAD

Setup: Create an Autodesk educational account or download the open-source FreeCAD.

Design your custom chassis, motor mounts, and enclosures for 3D printing.

⬇ Get CAD Software

🧠

Ultralytics YOLO

Setup: pip install ultralytics. Requires PyTorch.

Diagram & Sourcing

Draw a schematic. Gather your Arduino Uno, sensors, jumper wires, and breadboard.

Breadboard Prototyping

Connect everything on a breadboard without soldering. Double-check polarity and power limits.

IDE Setup & Libraries

Open Arduino IDE, select your board/COM port, and install required libraries via Library Manager.

Coding & Uploading

Write your setup() and loop(). Compile and flash the code via USB. Debug using Serial Monitor.

Soldering & Assembly

Transfer the verified circuit from breadboard to a perfboard/PCB and mount inside a custom 3D printed case.

Embedded Toolkit

Arduino Tools & Setup

The software and hardware essentials you need to start building Arduino projects.

🔌

Arduino IDE 2.x

Setup: Download installer, run, install included USB drivers.

The official development environment. Features autocompletion, board manager, and an integrated serial monitor.

⬇ Download IDE

🌐

Wokwi Simulator

Setup: Web-based. No installation required. Just sign up (optional).

Test your Arduino code and circuit wiring completely in the browser before buying physical hardware.

⬇ Open Wokwi

🎨

Fritzing

Setup: Download standard binary. (Paid, or compile from source for free).

Create visual breadboard wiring diagrams and convert them into PCB layouts for professional manufacturing.

⬇ Get Fritzing

⚡

CH340 USB Driver

Setup: Download and run the CH340 installer (Windows/Mac).

Many cheap Arduino clones use the CH340 serial chip instead of FTDI. You need this driver for your PC to recognize the board.

⬇ Download Driver

📏

Digital Multimeter

Hardware Requirement. Get one with continuity beeper.

Essential for checking voltages, measuring resistance, and finding short circuits in your wiring.

🛒 View Options

🔥

Soldering Iron Kit

Hardware Requirement. Recommendation: Pinecil or Hakko FX-888D.

Move beyond breadboards. Solder components to perfboards for permanent, reliable electrical connections.

🛒 Get a Pinecil

Learning hub · library

A.K. Gupta

Automation & Electronics

The Art of Electronics

Horowitz & Hill

Automation & Electronics

Community challenges

Test what you've built against others

Seasonal robotics challenges with a live leaderboard and shareable badges.

Line Follower Sprint · ends in 6 days

Rank	Builder	Project
1Aarav M.	Line Follower Robot	4.2s lap
2Priya S.	Line Follower Robot	4.6s lap
3Devon K.	Line Follower Robot	4.9s lap
4Marisol T.	Line Follower Robot	5.1s lap
5Kenji H.	Line Follower Robot	5.3s lap

Your progress

A dashboard that tracks the whole build

Skills, projects, simulations, and certificates — one panel, always current.

Robotics dashboard

Skill Tracker

Robotics

AI / CV

Embedded

ROS 2

Project Portfolio

projects completed

🛞

👁️

📦

Book Library

of 25 summaries read

Recent simulation runs

Obstacle Avoidance — GazeboPassed

SLAM Loop Closure — WebotsDrift detected

Pick-and-Place — CoppeliaSimPassed

Quadruped Gait Test — CoppeliaSimUnstable at 1.2x speed

Uptime this week

14.2h

across 3 simulation environments

Earned

Robotics Foundations · Embedded Systems Engineer

In progress

Computer Vision Engineer — 70% complete

Certificates

🏅

🏆

Recent AI Interactions

Debugged Arduino PWM codeResolved

Explained Kalman Filter mathResolved

ROS 2 Launch file errorPending

AI Insights

You tend to struggle with C++ pointers. Recommended review: Memory Management Module.

Start with node 01.

Mathematics & Physics is the first step on the trace — six modules to your first working robot.

Begin the roadmap →