The Elephant AI 3D Vision Kit is a beginner-friendly, all-in-one AI training platform designed for robotic arm learning, object recognition, and intelligent automation. Compatible with myCobot 320 (M5 & PI) models, it offers integrated positioning, automatic sorting, and gripping algorithms via 5 advanced visual recognition models, 7 practical application scenarios, and versatile end effector support.
Key Features
-
Includes 5 major visual recognition algorithms such as YOLOv5 Recognition, Object Detection, Color and Shape Recognition, and AR Code Processing.
-
Enables 7 application scenarios like intelligent gripping, feature point recognition, and real-world object classification using 2D and 3D relationships.
-
Comes with 2 gripping modules – the myGripper F100 and the myCobot Pro Suction Pump, perfect for a range of robotic operations.
Compatibility
This kit is compatible with 6 robotic arm models from Elephant Robotics, supporting both M5Stack and Raspberry Pi interfaces. This makes it ideal for users seeking flexible robotic development in both software and hardware environments.
What You’ll Learn
Use the Vision Education Kit to explore:
-
OpenCV, Python, and ROS programming
-
Inverse kinematics, robot vision, and hand-eye calibration
-
Robot control principles and end-effector integration
Whether you're building basic workflows or advanced AI recognition paths, this kit provides the necessary tools to level up your robotics expertise.
What’s in the Box
The kit includes:
-
Colored wooden blocks
-
Fruit model set
-
Shape and classification cards
-
QR and AR codes
-
Camera + mounting bracket
-
Cables, brackets, and hardware
-
Baseboard for myCobot 320
Real-World Use Cases
See how this kit is already transforming classrooms and innovation labs:
Video Learning Included
Get started fast with official tutorials:
-
myCobot 320 AI Kit Unboxing
-
How to Use myGripper F100
-
Firmware Setup and Configuration
Technical Comparison: PI vs M5
The myCobot 320 AI Kit is available in two variants:
Raspberry Pi Version (PI) and ESP32 Version (M5). Key differences:
-
PI: 64-bit CPU, 2G RAM, OpenCV support, Dual WiFi + Bluetooth
-
M5: Lightweight control via ESP32, ideal for embedded environments
| Feature | myCobot 320 PI | myCobot 320 M5 |
|---|---|---|
| Degree of Freedom | 6 | 6 |
| Payload | 1000g | 1000g |
| Working Radius | 350mm | 350mm |
| Positioning Accuracy | ±0.5mm | ±0.5mm |
| Weight | 3360g | 3360g |
| Working Temperature | -5℃ ~ 45℃ | -5℃ ~ 45℃ |
| Working Lifespan | 2000h | 2000h |
| Power Input | DC 24V 120W | DC 24V 120W |
| Joint Maximum Speed | 160°/s | 160°/s |
| MASTER | RASPBERRY PI | ESP32 |
| SOC | Broadcom BCM2711 | / |
| CPU | 64bit 1.5GHz 4B | / |
| GPU | 500 MHz VideoCore VI | / |
| RAM | 2G | / |
| Bluetooth | Dual mode Bluetooth 2.4G/5G | Dual mode Bluetooth 2.4G/5G |
| WiFi | 802.11AC, Dual model WiFi | 802.11ac |
| AUXILIARY CONTROL | ESP32 | / |
| Computing Performance | 600DMIPS | 600DMIPS |
| SRAM | 520KB | 520KB |
| Flash | 4M | 4M |
| USB Ports | USB3.0 ×2, USB2.0 ×2 | / |
| Type-C | 1 | / |
| Network Port | Gigabit Ethernet Port ×1 | / |
| Audio | 3.5mm ×1 | / |
| HDMI | microHDMI ×2 | / |
| Development Platforms (OS) | Windows, Linux, MacOS | Windows, Linux, MacOS |
| Programming Languages | Python | Python, C++, C#, JavaScript |
| ROS Version | ROS1 | ROS1 |
| Communication Protocols | Serial Communication, MODBUS, TCP/IP | Serial Communication, MODBUS, TCP/IP |
| Development Software | myBlockly, myStudio | myBlockly, myStudio, UiFlow, Arduino |
| I/O Interfaces | INPUT ×6, OUTPUT ×6 | INPUT ×6, OUTPUT ×6 |
Why Choose This Kit?
-
Comprehensive AI + Robotics Learning
-
Modular, Expandable, and Open Source
-
High Compatibility & Powerful Software Stack
-
Ideal for Classrooms, Labs, and DIY Developers