Arduino-compatible, ESP32-controlled, 2-wheeled Balancing Robot
The Elektor Mini-Wheelie is an experimental autonomous self-balancing robot platform. Based on an ESP32-S3 microcontroller, the self-balancing robot is fully programmable using the Arduino environment and open-source libraries. Its wireless capabilities allow it to be controlled remotely over Wi-Fi, Bluetooth or ESP-NOW or to communicate with a user or even another robot.
An ultrasonic transducer is available for detecting obstacles. Its color display can be used for displaying cute facial expressions or, for the more down-to-earth users, cryptic debug messages.
The robot comes as a neat kit of parts that you must assemble yourself. Everything is included, even a screwdriver.
Note: The Mini-Wheelie is an educational development platform intended for learning, experimentation, and robotics development. It is not classified as a toy for children, and its features, documentation, and intended audience reflect this purpose. The product is aimed at students, educators, and developers who wish to explore robotics, programming, and hardware integration in an educational setting.
Specifications
ESP32-S3 microcontroller with Wi-Fi and Bluetooth
MPU6050 6-axis Inertial Measurement Unit (IMU)
Two independently controlled 12 V electric motors with tachometer
Ultrasonic transducer
2.9" TFT color display (320 x 240)
MicroSD card slot
Battery power monitor
3S rechargeable Li-Po battery (11.1 V/2200 mAh)
Battery charger included
Arduino-based open-source software
Dimensions (W x L x H): 23 x 8 x 13 cm
Included
1x ESP32-S3 Mainboard + MPU6050 module
1x LCD board (2.9 inch)
1x Ultrasonic sensor
1x Battery pack (2200 mAh)
1x Battery charger
1x Motor tyre kit
1x Case board
1x Acrylic board
1x Screwdriver
1x Protective strip
1x Flex cable B (8 cm)
1x Flex cable A (12 cm)
1x Flex cable C
4x Copper column A (25 mm)
4x Copper column B (55 mm)
4x Copper column C (5 mm)
2x Plastic nylon column
8x Screws A (10 mm)
24x Screws B (M3x5)
8x Nuts
24x Metal washers
2x Zip tie
1x MicroSD card (32 GB)
Downloads
Documentation
For a limited time, the Joy-Pi Advanced is available in a great-value bundle with a Raspberry Pi 4 (8 GB)!
The Joy-Pi Advanced is a compact and powerful device that allows you to realize your projects quickly and easily. Whether you already have a lot of experience, or next to none, the Joy-Pi Advanced lets you unleash your creativity. Thanks to its compatibility with a wide range of platforms, including Raspberry Pi, Raspberry Pi Pico, Arduino Nano, BBC micro:bit, and NodeMCU ESP32, you can easily and quickly access your preferred platform.
In addition, the Joy-Pi Advanced features more than 30 stations, lessons, and modules, giving you an unlimited variety of ways to get your projects done. With the self-developed learning center, you can not only improve your skills but also create new projects. The learning center offers a wealth of information and tutorials that will guide you step by step through your projects.
Joy-Pi Advanced is characterized in particular by its intelligent switch units, which allow an extended use of the available pins. A total of three switch units are integrated, each equipped with 12 individual switches that provide precise control of the connected sensors and modules. This system solves the well-known problem of limited pin count that occurs with conventional microcontrollers. The switch units allow you to operate a large number of sensors and modules in parallel by switching them on and off individually. This simulates multiple pin assignment, allowing you to exploit the full power of your projects without compromising functionality.
By combining innovative adapter boards and the micro:bit slot, you can achieve seamless compatibility with a wide range of microcontrollers such as Raspberry Pi Pico, NodeMCU ESP32, micro:mit and Arduino Nano. The specially developed adapter boards are designed to perfectly match the respective microcontroller. By plugging the microcontroller onto the appropriate adapter board and then plugging it into the micro:bit slot, the Joy-Pi Advanced quickly and easily becomes compatible with the different microcontrollers. This allows seamless integration of your preferred platform and the ability to combine the strengths of the different microcontrollers in your projects. This way, you can fully focus on your creative projects without worrying about the compatibility of different microcontrollers. The Joy-Pi Advanced simplifies the development process and gives you the possibility to design your projects flexibly and individually.
Features
Highly integrated development platform & learning center
Fast, easy & wireless combination of various sensors & actuators
Installation option for Raspberry Pi 4
Compatible with various microcontrollers
Self-developed, didactic learning platform for Raspberry Pi & Windows
Specifications
Compatible to
Raspberry Pi 4, Arduino Nano, NodeMCU ESP32, BBC micro:bit, Raspberry Pi Pico
Installed sensors, actuators & components
39
Learning platform
Over 40 entries in the know-ledge database, 10 projects, 10 learning tasks, 14 visions
Displays
7-segment display, 16x2 display, 1.8“ TFT display, 0.96" OLED display, 8x8 RGB matrix
Sensors
DS18B20, shock sensor, hall sensor, barometer, sound sensor, gyroscope, PIR sensor, Light barrier, NTC, Light sensor, 6x touch sensor, color sensor, ultrasonic distance sensor, DHT11 temperature & humidity sensor
Control
Joystick, 5x switches, potentiometer, rotary encoder, 4x4 button matrix, relays, PWM fan
Motors
Servo interface, Stepper motor interface, Vibration motor
Measuring & conversion modules
Analog-Digital Converter, Level converter, voltmeter, Variable voltage supply
Other components
RTC real time clock, buzzer, EEPROM memory, infrared receiver, breadboard, RFID reader
Adapter boards
Adapter for NodeMCU ESP32, Arduino Nano & Raspberry Pi Pico, Board connectors for Raspberry Pi & External Boards
Electronic components
Infrared remote control, RFID chip, RFID card, 6x alligator clips, microSD card reader, servo motor, stepper motor, 32 GB microSD card
Components
40x resistors, 3x green LEDs, 3x yellow LEDs, 3x red LEDs, 1x transistor, 5x buttons, 1x potentiometer, 2x capacitors
Other accessories
Screw assortment, screwdriver, accessory storage bag, power supply & power cable, servo mount
Power supply
Built-in power supply: 36 W, 12 V, 3 A Case connector: Small device plug C8
Voltage outputs
12 V, 5 V, 3.3 V, Variable voltage output (2-11 V)
Data buses & signal outputs
I²C, SPI, Analog to digital converter
Battery (RTC)
CR2032
Dimensions
327 x 200 x 52 mm
Included
Raspberry Pi 4 (8 GB RAM)
Downloads
Joy-Pi website
Datasheet
Manual
Comprehensive Book-Hardware Bundle for the RP2040 Microcontroller with over 80 Projects
Unlock the potential of modern controller technology with the Raspberry Pi Pico in this bundle. Perfect for both beginners and experienced users, the easy-to-follow guide takes you from the basics of electronics to the complexities of digital signal processing. With the Raspberry Pi Pico, the dedicated hardware kit and MicroPython programming, you will learn the key principles of circuit design, data collection, and processing.
Get hands-on with over 80 projects like a stopwatch with an OLED display, a laser distance meter, and a servo-controlled fan. These projects are designed to help you apply what you've learned in real-world scenarios. The book also covers advanced topics like wireless RFID technology, object detection, and sensor integration for robotics.
Whether you're looking to build your skills in electronics or dive deeper into embedded systems, this bundle is the perfect resource to help you explore the full potential of the Raspberry Pi Pico.
Contents of the Bundle
1x Project Book (273 pages)
1x Smart Car Kit
Electronic Parts
2x Solderless breadboard (400 holes)
1x Solderless breadboard (170 holes)
5x Colorful 5 mm LEDs (green, red, blue, yellow and white)
1x Laser transmitter
1x Passive buzzer
1x Micro USB cable (30 cm)
1x 65 Jumper wires
1x 20 cm male to female Dupont wire
1x Clear case
1x Magnet (diameter: 8 mm, thickness: 5 mm)
1x Rotary potentiometer
10x 2 KΩ resistors
2x M2.5x30 mm copper pillars
10x Phillips pan head screws
10x M2.5 nickel hex nuts
1x 2-inch dual-purpose screwdriver
Modules
1x RGB module
1x 9G servo
1x Dual-axis XY joystick module
1x RC522 RFID module
1x 4 Bits digital LED display module
1x Traffic light display module
1x Rotary Encoder module
1x 1602 LCD Display module (Blue)
1x Photoresistor module
1x DC motor with male Dupont wire
1x Fan blade
1x Raindrops module
1x OLED module
1x Membrane switch keyboard
1x Mini magnetic spring module
1x Infrared remote control
1x Infrared receiver module
1x DC stepper motor driver board
1x Button
Sensors
1x Vibration sensor
1x Soil moisture sensor
1x Sound sensor
1x Mini PIR motion sensor
1x Temperature & Humidity sensor
1x Flame sensor
2x Crash sensor
2x Tracking sensor
1x Ultrasonic sensor
This bundle contains the AI Fusion Lab Kit – which includes over 450 components – and the book Physical AI with Raspberry Pi, offering a comprehensive, hands-on introduction to AI, electronics, and Python programming.
Build practical projects that combine computer vision, voice interaction, sensors, motors, and large language models. The included book guides you step by step from Raspberry Pi setup and hardware control to advanced applications using OpenCV, MediaPipe, YOLO, OpenAI services, and local Ollama models.
SunFounder AI Fusion Lab Kit for Raspberry Pi (450+ parts)
The AI Fusion Lab Kit transforms a Raspberry Pi into a complete platform for learning artificial intelligence, electronics, and Python programming. It supports Raspberry Pi 5, 4B, 3B+, 3B, and Zero 2 W, as well as popular AI models and services including ChatGPT, Gemini, Grok, DeepSeek, Qwen, Doubao, and Ollama.
With its camera, Pan-Tilt HAT, 10-axis sensor module, microphone, speaker, and electronic components, the kit enables users to build interactive AI projects that combine voice, vision, movement, and sensor data. It is suitable for beginners, students, makers, and anyone who wants to explore practical AI applications.
Build Multimodal AI Projects
Combine large language models with speech recognition, text-to-speech, camera vision, sensors, and breadboard circuits. Create projects that can listen, speak, see, react, and interact with their environment while learning how modern AI systems work.
Explore AI Vision and Object Tracking
Use YOLO, OpenCV, and MediaPipe to detect and track objects, colors, faces, and human movement. The motorized Pan-Tilt HAT allows the camera to follow targets automatically, demonstrating how computer vision and motion control work together.
Voice AI with the Fusion HAT+
The Fusion HAT+ simplifies hardware control through a unified Python library and supports safe Raspberry Pi shutdown. Its integrated microphone and speaker make it easy to develop voice-controlled assistants and other conversational AI applications.
Compatible with Raspberry Pi 5, 4B, 3B+, 3B, and Zero 2 W
Supports multiple cloud-based and local AI models
Includes camera, Pan-Tilt HAT, 10DOF module, microphone, and speaker
Projects covering voice AI, computer vision, sensors, and electronics
Video lessons, documentation, and sample code
Book: Physical AI with Raspberry Pi
Using the Raspberry Pi and the AI Fusion Lab Kit, this book takes you step by step from the fundamentals of hardware interfacing to the development of intelligent systems capable of seeing, listening, speaking, and making decisions. Whether you are a student, educator, maker, or professional engineer, you will gain valuable experience through projects designed to build both knowledge and confidence.
Beginning with system setup, GPIO programming, sensors, displays, motors, and automation techniques, the book gradually introduces more advanced topics, including speech recognition, text-to-speech synthesis, conversational AI, object recognition, and machine learning. Along the way, you will learn how to create voice-controlled devices, intelligent assistants, camera tracking systems, environmental monitoring solutions, and AI-powered embedded applications.
Explore modern technologies such as OpenAI services, Ollama local large language models, OpenCV, MediaPipe, and YOLO object detection while gaining practical experience with real-world applications used in robotics, industrial automation, smart devices, and autonomous systems. You’ll discover how intelligent software can interact with sensors, cameras, motors, and displays to create systems that respond naturally to the world around them.
Rather than simply following code examples, you will be encouraged to experiment, customize, and expand every project. The techniques and skills developed throughout this book provide a solid foundation for future innovation, rapid prototyping, research, and product development. If you are ready to move beyond traditional electronics and explore the future of intelligent embedded systems, this book will guide you every step of the way.
This bundle contains:
SunFounder AI Fusion Lab Kit for Raspberry Pi (normal price: €100)
Book: Physical AI with Raspberry Pi (normal price: €45)
Not included
Raspberry Pi
Over 180 Projects with Raspberry Pi, Pico W, Arduino, and ESP32
This bundle contains the Universal Maker Sensor Kit, which consists of many sensors, actuators, displays, and motors. It’s perfect for environmental monitoring, smart home projects, robotics, and game controllers.
The new Elektor book describes the design of many projects using the kit together with the popular Raspberry Pi, Raspberry Pi Pico W, Arduino Uno, and the ESP32 family of development boards. You can choose any of these development boards for your projects and either use the provided programs as they are, or modify these programs to suit your applications.
This bundle contains:
Book: Universal Maker Sensor Kit (normal price: €45)
Universal Maker Sensor Kit for Raspberry Pi, Pico W, Arduino, ESP32 (normal price: €70)
Book: Universal Maker Sensor Kit
Learn to use more than 35 Sensors and Actuators with C++, Python, and MicroPython
This book contains over 180 projects for all four major development boards (Arduino, Raspberry Pi, Pico W, and ESP32). Depending on the development board, projects are available in the C, Python, or MicroPython programming languages.
The project titles, brief descriptions, wiring diagrams, and full program listings together with their detailed descriptions are given in the guide.
Universal Maker Sensor Kit (for Raspberry Pi, Pico W, Arduino, ESP32)
Discover endless creativity with the Universal Maker Sensor Kit, designed for use with Raspberry Pi, Pico W, Arduino, and ESP32. This versatile kit offers compatibility across popular development platforms, including Arduino Uno R4 Minima/WiFi, Uno R3, Mega 2560, Raspberry Pi 5, 4, 3B+, 3B, Zero, Pico W, and ESP32.
Featuring over 35 sensors, actuators, and displays, it's perfect for projects ranging from environmental monitoring and smart home automation to robotics and interactive gaming. Step-by-step tutorials in C/C++, Python, and MicroPython guide beginners and experienced makers alike through 169 exciting projects.
Features
Wide Compatibility: Fully supports Arduino (Uno R3, Uno R4 Minima/WiFi, Mega 2560), Raspberry Pi (5, 4, 3B+, 3B, Zero, Pico W), and ESP32, enabling extensive flexibility across numerous development platforms. Includes instructions for building 169 projects.
Comprehensive Components: Features more than 35 sensors, actuators, and display modules suitable for diverse projects such as environmental monitoring, smart home automation, robotics, and interactive game controllers.
Detailed Tutorials: Provides clear, step-by-step tutorials covering Arduino, Raspberry Pi, Pico W, ESP32, and each included component. Tutorials are available in C/C++, Python, and MicroPython, catering effectively to both beginners and experienced makers.
Suitable for All Skill Levels: Offers structured projects designed to guide users seamlessly from beginner to advanced proficiency in electronics and programming, enhancing creativity and technical expertise.
Kit includes
Breadboard
Button Module
Capacitive Soil Moisture Module
Flame Sensor Module
Gas/Smoke Sensor Module (MQ2)
Gyroscope & Accelerometer Module (MPU6050)
Hall Sensor Module
Infrared Speed Sensor Module
IR Obstacle Avoidance Sensor Module
Joystick Module
PCF8591 ADC DAC Converter Module
Photoresistor Module
PIR Motion Module (HC-SR501)
Potentiometer Module
Pulse Oximeter and Heart Rate Sensor Module (MAX30102)
Raindrop Detection Module
Real Time Clock Module (DS1302)
Rotary Encoder Module
Temperature Sensor Module (DS18B20)
Temperature and Humidity Sensor Module (DHT11)
Temperature, Humidity & Pressure Sensor (BMP280)
Time of Flight Micro-LIDAR Distance Sensor (VL53L0X)
Touch Sensor Module
Ultrasonic Sensor Module (HC-SR04)
Vibration Sensor Module (SW-420)
Water Level Sensor Module
I²C LCD 1602
OLED Display Module (SSD1306)
RGB LED Module
Traffic Light Module
5 V Relay Module
Centrifugal Pump
L9110 Motor Driver Module
Passive Buzzer Module
Servo Motor (SG90)
TT Motor
ESP8266 Module
JDY-31 Bluetooth Module
Power Supply Module
Documentation
Online Tutorial
Temporary Delay in the Delivery of Unitree Robots
Like many other suppliers, we are currently experiencing delays in the delivery of Unitree robots. A shipment from our supplier is currently held in customs, which has unfortunately led to later-than-planned deliveries for previously placed orders. We are actively working with our supplier to resolve this issue and expect more clarity soon, but at this time, we cannot provide any guarantees.
Additionally, a new shipment is already on its way, though it will take some time to arrive. Since other suppliers are facing similar challenges, switching to a different provider is unlikely to result in a faster solution. Our top priority remains fulfilling existing orders.
If you have any questions or would like to update your order, please do not hesitate to contact our customer service team. We will keep you informed of any further developments.
Unitree Go2 series consists of quadruped robots for the research & development of autonomous systems in the fields of human-robot interaction (HRI), SLAM & transportation. Due to the four legs, as well as the 12DOF, this robot can handle a variety of different terrains. The Go2 comes with a perfected drive & power management system, which enables a speed (depending on the version) of up to 3.7 m/s or 11.88 km/h with an operating time of up to 4 hours. Furthermore, the motors have a torque of 45 N.m at the body/thighs and at the knees, which also allow jumps or backflips.
Features
Super Recognition System: 4D LIDAR L1
Max Running Speed: approx. 5 m/s
Peak Joint Torque: approx. 45 N.m
Wireless Module: WiFi 6/Bluetooth/4G
Ultra-long battery Endurance: approx. 2-4 h (long battery life measured in real life)
Intelligent Side-follow System: ISS 2.0
Specifications
Tracking module: Remote-controlled or automatic tracking
Front camera: Image tansmission Resolution 1280x720, FOV 120°, Ultra wide angle lens deliver rich clarity
Front lamp: Brightly lights the way ahead
4D LiDAR L1: 360°x90° omnidirectional ultra-wide-angle scanning allows automatic avoidance with small blind spot and stable operation
12 knee joint motors: Strong and powerful, Beautiful and simple, Brandy new visual experience
Intercom microphone: Effective communication with no scenario restrictions
Self-retracting strap: Easy to carry and load things
More stable, more powerful with advanced devices: 3D LiDAR, 4G ESIM Card, WiFi 6 with Dual-band, Bluetooth 5.2 for stable connection and remote control
Powerful Computing Core: Motion controller, High-performance ARM processor, Improved Al algorithm processor, External ORIN NX/NANO
Smart battery: Standard 8000 mAh battery, Long-endurance 15000 mAh battery, Protection from over-temp, overcharge and short-circuit
Speaker for music play: Listen to music as your pleasure
Unitree Go2 Variants
The Go2 impresses not only with its technical capabilities, but also with a modern and slim design that gives it a futuristic look and makes it a real eye-catcher. The Go2 Air is specially designed for demos and presentations. With its basic features, it offers a solid basis for demonstrating the movement capabilities and functionality of a four-legged robot. Important: The Go2 Air is delivered without a controller. This can be purchased optionally.
With a powerful 8-core high-performance CPU, the Pro and Edu offer impressive computing power required for complex tasks and demanding calculations. This enables faster and more efficient data processing and makes the Pro and Edu a reliable partner for your projects.
From the Edu version onwards, the Go2 is programmable and opens up endless possibilities for developing and researching your own robotics applications. The Go2 is also able to handle a step height of up to 14 cm. This makes it an ideal tool for research, education and entry into the world of robotics.
The Go2 Edu comes with a remote controller that gives you easy and intuitive control. You also get a docking station with impressive computing power of 100 TOPS, which is equipped with powerful AI algorithms and offers you technical support.
Go2 Edu is equipped with a powerful 15000 mAh battery that gives it an impressive runtime of up to 4 hours. This long operating time allows the robot to carry out longer exploration missions and complete demanding tasks.
Model Comparison
Air
Pro
Edu/Edu Plus
Dimensions (standing)
70 x 31 x 40 cm
70 x 31 x 40 cm
70 x 31 x 40 cm
Dimensions (crouching)
76 x 31 x 20 cm
76 x 31 x 20 cm
76 x 31 x 20 cm
Material
Aluminium alloy + High strength engineering plastic
Aluminium alloy + High strength engineering plastic
Aluminium alloy + High strength engineering plastic
Weight (with battery)
about 15 kg
about 15 kg
about 15 kg
Voltage
28~33.6 V
28~33.6 V
28~33.6 V
Peaking capacity
about 3000 W
about 3000 W
about 3000 W
Payload
≈7 kg (MAX ~ 10 kg)
≈8 kg (MAX ~ 10 kg)
≈8 kg (MAX ~ 12 kg)
Speed
0~2.5 m/s
0~3.5 m/s
0~3.7 m/s (MAX ~ 5 m/s)
Max Climb Drop Height
about 15 cm
about 16 cm
about 16 cm
Max Climb Angle
30°
40°
40°
Basic Computing Power
N/A
8-core High-performance CPU
8-core High-performance CPU
Aluminum knee joint motor
12 set
12 set
12 set
Intra-joint circuit (knee)
✓
✓
✓
Joint Heat Pipe Cooler
✓
✓
✓
Range of Motion
Body: −48~48°
Body: −48~48°
Body: −48~48°
Thigh: −200°~90°
Thigh: −200°~90°
Thigh: −200°~90°
Shank: −156°~−48°
Shank: −156°~−48°
Shank: −156°~−48°
Max Torque
N/A
about 45 N.m
about 45 N.m
Super-wide-angle 3D LiDAR
✓
✓
✓
Wireless Vector Positioning Tracking Module
N/A
✓
✓
HD Wide-angle Camera
✓
✓
✓
Foot-end force sensor
N/A
N/A
✓
Basic Action
✓
✓
✓
Auto-scaling strap
N/A
✓
N/A
Upgraded Intelligent OTA
✓
✓
✓
RTT 2.0 Image Transmission
✓
✓
✓
App Basic Remote Control
✓
✓
✓
App Data Viewing
✓
✓
✓
App Graphical Programme
✓
✓
✓
Front Lamp (3 W)
✓
✓
✓
WiFi 6 with Dual-band
✓
✓
✓
Bluetooth 5.2/4.2/2.1
✓
✓
✓
4G Module
N/A
CN/GB
CN/GB
Voice Function
N/A
✓
✓
Music Playback
N/A
✓
✓
ISS 2.0 Intelligent side-follow system
N/A
✓
✓
Intelligent detection and avoidance
✓
✓
✓
Secondary development
N/A
N/A
✓
Manual controller
Optional
Optional
✓
High computing power module
N/A
N/A
Edu: 40 TOPS computing power
Edu Plus: 100 TOPS computing power
NVIDIA Jetson Orin (optional)
Smart Battery
Standard (8000 mAh)
Standard (8000 mAh)
Long endurance (15000 mAh)
Battery Life
1-2 h
1-2 h
2-4 h
Charger
Standard (33.6 V, 3.5 A)
Standard (33.6 V, 3.5 A)
Fast charge (33.6 V, 9 A)
Included
1x Unitree Go2 Pro
1x Unitree Go2 Battery (8000 mAh)
Downloads
Documentation
iOS/Android apps
GitHub
This hands-on bundle lets you build real Edge AI applications with the Raspberry Pi
Raspberry Pi AI HAT+ (13 TOPS)
The Raspberry Pi AI HAT+ (13 TOPS) is an expansion board designed for the Raspberry Pi 5, featuring an integrated Hailo AI accelerator. This add-on offers a cost-effective, efficient, and accessible approach to incorporating high-performance AI capabilities, with applications spanning process control, security, home automation, and robotics.
The AI HAT+ connects via the Raspberry Pi 5’s PCIe Gen3 interface. When the Raspberry Pi 5 is running a current version of the Raspberry Pi OS, it automatically detects the onboard Hailo accelerator, making the neural processing unit (NPU) available for AI tasks. Additionally, the rpicam-apps camera applications included in Raspberry Pi OS seamlessly support the AI module, automatically using the NPU for compatible post-processing functions.
Raspberry Pi Camera Module 3
Raspberry Pi Camera Module 3 is a compact camera from Raspberry Pi. It offers an IMX708 12-megapixel sensor with HDR, and features phase detection autofocus. Camera Module 3 is available in standard and wide-angle variants, both of which are available with or without an infrared cut filter.
Camera Module 3 can be used to take full HD video as well as stills photographs, and features an HDR mode up to 3 megapixels. Its operation is fully supported by the libcamera library, including Camera Module 3’s rapid autofocus feature: this makes it easy for beginners to use, while offering plenty for advanced users. Camera Module 3 is compatible with all Raspberry Pi computers.
Book: Edge AI Made Practical – AI Projects for the Raspberry Pi with the AI HAT+
Edge AI is transforming everyday devices by putting intelligence where it matters most: directly inside the hardware. With on-device inference, a camera can recognize a visitor instantly, a phone can translate speech without streaming audio to the cloud, and a wearable can detect anomalies in real time—fast, private, and reliable even when the network disappears.
This book is your practical guide to building exactly those kinds of systems with the Raspberry Pi AI HAT+ and the Hailo-8L accelerator. You’ll start with clear foundations: core AI and machine-learning concepts, how neural networks work, and what truly distinguishes Edge AI from cloud AI—plus an honest look at ethical considerations and future impacts.
This bundle contains:
Book: Edge AI Made Practical
Raspberry Pi AI HAT+ (13 TOPS)
Raspberry Pi Camera Module 3
Active Cooler for Raspberry Pi 5
FPC Display Cable for Raspberry Pi 5 (300 mm)
Elektor Component Kit
40-pin GPIO Header
Traffic Light Module
LED red 5 V with built-in resistor
LED yellow 5 V with built-in resistor
LED green 5 V with built-in resistor
LED blue 5 V with built-in resistor
Breadboard (400 Tie-points)
10 Dupont wires male-female
DHT22 sensor module
Servo motor
Required
Raspberry Pi 5
Temporary Delay in the Delivery of Unitree Robots
Like many other suppliers, we are currently experiencing delays in the delivery of Unitree robots. A shipment from our supplier is currently held in customs, which has unfortunately led to later-than-planned deliveries for previously placed orders. We are actively working with our supplier to resolve this issue and expect more clarity soon, but at this time, we cannot provide any guarantees.
Additionally, a new shipment is already on its way, though it will take some time to arrive. Since other suppliers are facing similar challenges, switching to a different provider is unlikely to result in a faster solution. Our top priority remains fulfilling existing orders.
If you have any questions or would like to update your order, please do not hesitate to contact our customer service team. We will keep you informed of any further developments.
Unitree Go2 series consists of quadruped robots for the research & development of autonomous systems in the fields of human-robot interaction (HRI), SLAM & transportation. Due to the four legs, as well as the 12DOF, this robot can handle a variety of different terrains. The Go2 comes with a perfected drive & power management system, which enables a speed (depending on the version) of up to 3.7 m/s or 11.88 km/h with an operating time of up to 4 hours. Furthermore, the motors have a torque of 45 N.m at the body/thighs and at the knees, which also allow jumps or backflips.
Features
Super Recognition System: 4D LIDAR L1
Max Running Speed: approx. 5 m/s
Peak Joint Torque: approx. 45 N.m
Wireless Module: WiFi 6/Bluetooth/4G
Ultra-long battery Endurance: approx. 2-4 h (long battery life measured in real life)
Intelligent Side-follow System: ISS 2.0
Specifications
Tracking module: Remote-controlled or automatic tracking
Front camera: Image tansmission Resolution 1280x720, FOV 120°, Ultra wide angle lens deliver rich clarity
Front lamp: Brightly lights the way ahead
4D LiDAR L1: 360°x90° omnidirectional ultra-wide-angle scanning allows automatic avoidance with small blind spot and stable operation
12 knee joint motors: Strong and powerful, Beautiful and simple, Brandy new visual experience
Intercom microphone: Effective communication with no scenario restrictions
Self-retracting strap: Easy to carry and load things
More stable, more powerful with advanced devices: 3D LiDAR, 4G ESIM Card, WiFi 6 with Dual-band, Bluetooth 5.2 for stable connection and remote control
Powerful Computing Core: Motion controller, High-performance ARM processor, Improved Al algorithm processor, External ORIN NX/NANO
Smart battery: Standard 8000 mAh battery, Long-endurance 15000 mAh battery, Protection from over-temp, overcharge and short-circuit
Speaker for music play: Listen to music as your pleasure
Unitree Go2 Variants
The Go2 impresses not only with its technical capabilities, but also with a modern and slim design that gives it a futuristic look and makes it a real eye-catcher. The Go2 Air is specially designed for demos and presentations. With its basic features, it offers a solid basis for demonstrating the movement capabilities and functionality of a four-legged robot. Important: The Go2 Air is delivered without a controller. This can be purchased optionally.
With a powerful 8-core high-performance CPU, the Pro and Edu offer impressive computing power required for complex tasks and demanding calculations. This enables faster and more efficient data processing and makes the Pro and Edu a reliable partner for your projects.
From the Edu version onwards, the Go2 is programmable and opens up endless possibilities for developing and researching your own robotics applications. The Go2 is also able to handle a step height of up to 14 cm. This makes it an ideal tool for research, education and entry into the world of robotics.
The Go2 Edu comes with a remote controller that gives you easy and intuitive control. You also get a docking station with impressive computing power of 100 TOPS, which is equipped with powerful AI algorithms and offers you technical support.
Go2 Edu is equipped with a powerful 15000 mAh battery that gives it an impressive runtime of up to 4 hours. This long operating time allows the robot to carry out longer exploration missions and complete demanding tasks.
Model Comparison
Air
Pro
Edu/Edu Plus
Dimensions (standing)
70 x 31 x 40 cm
70 x 31 x 40 cm
70 x 31 x 40 cm
Dimensions (crouching)
76 x 31 x 20 cm
76 x 31 x 20 cm
76 x 31 x 20 cm
Material
Aluminium alloy + High strength engineering plastic
Aluminium alloy + High strength engineering plastic
Aluminium alloy + High strength engineering plastic
Weight (with battery)
about 15 kg
about 15 kg
about 15 kg
Voltage
28~33.6 V
28~33.6 V
28~33.6 V
Peaking capacity
about 3000 W
about 3000 W
about 3000 W
Payload
≈7 kg (MAX ~ 10 kg)
≈8 kg (MAX ~ 10 kg)
≈8 kg (MAX ~ 12 kg)
Speed
0~2.5 m/s
0~3.5 m/s
0~3.7 m/s (MAX ~ 5 m/s)
Max Climb Drop Height
about 15 cm
about 16 cm
about 16 cm
Max Climb Angle
30°
40°
40°
Basic Computing Power
N/A
8-core High-performance CPU
8-core High-performance CPU
Aluminum knee joint motor
12 set
12 set
12 set
Intra-joint circuit (knee)
✓
✓
✓
Joint Heat Pipe Cooler
✓
✓
✓
Range of Motion
Body: −48~48°
Body: −48~48°
Body: −48~48°
Thigh: −200°~90°
Thigh: −200°~90°
Thigh: −200°~90°
Shank: −156°~−48°
Shank: −156°~−48°
Shank: −156°~−48°
Max Torque
N/A
about 45 N.m
about 45 N.m
Super-wide-angle 3D LiDAR
✓
✓
✓
Wireless Vector Positioning Tracking Module
N/A
✓
✓
HD Wide-angle Camera
✓
✓
✓
Foot-end force sensor
N/A
N/A
✓
Basic Action
✓
✓
✓
Auto-scaling strap
N/A
✓
N/A
Upgraded Intelligent OTA
✓
✓
✓
RTT 2.0 Image Transmission
✓
✓
✓
App Basic Remote Control
✓
✓
✓
App Data Viewing
✓
✓
✓
App Graphical Programme
✓
✓
✓
Front Lamp (3 W)
✓
✓
✓
WiFi 6 with Dual-band
✓
✓
✓
Bluetooth 5.2/4.2/2.1
✓
✓
✓
4G Module
N/A
CN/GB
CN/GB
Voice Function
N/A
✓
✓
Music Playback
N/A
✓
✓
ISS 2.0 Intelligent side-follow system
N/A
✓
✓
Intelligent detection and avoidance
✓
✓
✓
Secondary development
N/A
N/A
✓
Manual controller
Optional
Optional
✓
High computing power module
N/A
N/A
Edu: 40 TOPS computing power
Edu Plus: 100 TOPS computing power
NVIDIA Jetson Orin (optional)
Smart Battery
Standard (8000 mAh)
Standard (8000 mAh)
Long endurance (15000 mAh)
Battery Life
1-2 h
1-2 h
2-4 h
Charger
Standard (33.6 V, 3.5 A)
Standard (33.6 V, 3.5 A)
Fast charge (33.6 V, 9 A)
Included
1x Unitree Go2 Air
1x Unitree Go2 Battery (8000 mAh)
Downloads
Documentation
iOS/Android apps
GitHub
The Unitree G1 is a modern humanoid robot that impresses with its remarkable flexibility and advanced technology. With an exceptionally wide range of joint movement and up to 43 joint motors, it exceeds the agility of a typical human. Powered by imitation learning and reinforcement learning, its robotic systems are continuously developed and optimized through artificial intelligence.
One of the G1's most impressive features is its ability to autonomously move into a walking position as soon as it touches the ground – no external assistance required! It can immediately start moving, demonstrating a high level of independence and adaptability. The G1 is also equipped with a force-controlled, highly dexterous hand that operates with both sensitivity and precision, thanks to its combination of force and position control. This hand closely mimics human movements, allowing for precise object manipulation.
Features
Intel RealSense D435 Depth Camera
Livox MID-360 3D LiDAR
Microphone array (noise and echo cancellation)
5 W stereo speaker
Extra large quick release battery
Single arm degrees of freedom (shoulder 2 + elbow 2)
Hollow joint wiring of the whole machine (no external cables)
Maximum torque at joints 120 N.m
Single leg degrees of freedom (hip 3, knee 1, ankle 2)
Moving speed of 2 m/s
Specifications
Height, Width and Thickness (Stand)
1320 x 450 x 200 mm
Height, Width and Thickness (Fold)
690 x 450 x 300 mm
Weight (with Battery)
approx. 35 kg
Total Degrees of Freedom(Joint Freedom
23
Single Leg Degrees of Freedom
6
Waist Degrees of Freedom
1
Single Arm Degrees of Freedom
5
Joint output bearing
Industrial grade crossed roller bearings (high precision, high load capacity)
Joint motor
Low inertia high-speed internal rotor PMSM (Permanent Magnet Synchronous Motor – better response speed and heat dissipation)
Maximum Torque of Knee Joint
90 N.m
Arm Maximum Load
approx. 2 kg
Calf + Thigh Length
0.6 m
Arm Span
approx. 0.45 m
Extra Large Joint Movement Space
• Waist joint: Z ±155°• Knee joint: 0~165°• Hip joint: P ±154°, R -30~+170°, Y ±158°
Full Joint Hollow Electrical Routing
Yes
Joint Encoder
Dual encoder
Cooling System
Local air cooling
Power Supply
13 string Lithium battery
Basic Computing Power
8-core high-performance CPU
Sensing Sensor
Depth Camera + 3D LiDAR
Microphones
4 Microphone Array
Speaker
5 W stereo speaker
Wireless
WiFi 6, Bluetooth 5.2
Smart Battery (Quick Release)
9000 mAh
Charger
54 V/5 A
Manual Controller
Yes
Battery Life
approx. 2 hours
Upgraded Intelligent OTA
Yes
High-Performance LoRa Communication Device
The Elecrow ThinkNode M5 is a compact, portable LoRa communication device designed for off-grid messaging, location sharing, and outdoor use. Powered by an ESP32-S3 processor and an SX1262 LoRa chip, it comes pre-installed with Meshtastic firmware and can be configured conveniently via Bluetooth using the official Meshtastic app.
The integrated multi-GNSS module supports GPS, GLONASS, BeiDou, and QZSS for reliable positioning and location sharing. Messages and device information are displayed on the energy-efficient 1.54-inch e-paper screen, while the built-in 1,200 mAh rechargeable battery enables extended mobile operation.
With its integrated LoRa antenna, compact housing, RTC, and support for both Meshtastic and MeshCore firmware, the ThinkNode M5 is a practical communication solution for hiking, camping, remote locations, and emergency situations where conventional mobile networks are unavailable.
Features
LoRa communication
Multi-GNSS positioning
1,200 mAh rechargeable battery
1.54-inch e-paper display
Built-in RTC
Compatible with MeshCore
Compatible with Meshtastic
Specifications
Main Processor (ESP32-S3)
CPU/SoC
Xtensa 32-bit LX7 dual-core processor, clocked at up to 240 MHz
System Memory
512 KB SRAM, 8 MB PSRAM
Storage
4 MB Flash, 384 KB ROM
Firmware
Meshtastic firmware is fully compatible, and the signal is transmitted in LoRa mode
EPD (Electronic Paper Display)
Display Size
1.54 inch EPD (monochrome Ink screen)
Display Materials
E-lnk (Electronic Ink)
Resolution
200 x 200
Driver Chip
SSD1681 (via SPI or I²C interface)
Global Fresh Time
2s
Wireless Communication
Bluetooth
Bluetooth Low Energy and Bluetooth 5.0 (phone configuration)
LoRa
SX1262 LoRa Module, EU 868 MHz(external antenna)
Hardware
Interfaces
USB-C, RP-SMA
Function
GPS Location (GPS, GLONASS, BeiDou, QZSS), EPD Display, RTC, USB 2.0, PMU power management (built-in 1200 mAh lithium battery), buzzer, etc.
Button
Knob Switch, Function Button, Page Turn Button, GPS Switch, Reset Button
LED Indicator
Power supply, GPS/LoRa indication
Others
Power Input
5 V/1 A, supports USB or lithium battery power supply
Power consumption
The maximum working current is about 340 mA (CPU + LoRa module), and the low power consumption is about 34 uA
Operating Temperature
-10~50°C
Storage Temperature
-20~60°C
Relative humidity
10-95%, @ 40°C (non-condensing)
Enclosure
ABS Plastic
Dimensions
82 x 51.6 x 26.3 mm
Weight
81 g
Included
1x Elecrow ThinkNode M5
1x LoRa antenna
1x USB-C cable
1x Manual
Downloads
Manual
,
by Lobna Belarbi
Affordable Robot Kits to Kickstart Your Robotics Journey
Robotics is an exciting and rewarding field, but getting started can be intimidating—especially when it comes to choosing the right kit. Fortunately, Elektor offers a...
,
by Clemens Valens
Trying Out the Joy-it JT-PS1440-C 1.5 kW Power Supply (Review)
In today's high-powered world of e-bikes, electric scooters, and various other electronic vehicles, robust and adaptable power supplies are indispensable for motor testing and battery...