Build An AI Enabled Mobile Robot and Robotic Arm




Become an expert at modern robotics and get hands-on experience by building your own AI enabled robotic arm using Arduino and Raspberry Pi, and an advanced camera-enabled mobile robot. Gain experience in programming, electronics, 3D Printing, computer vision and much more. 

Meet Your New Opponent

Meet EvoArm, robot with the precision, accuracy, and intelligence to play real Tic-Tac-Toe against you, and win. And that’s just one of the things EvoArm can do. Refined with the input of hundreds of users, EvoArm is designed to give you the best deep dive into robotics from low-level electronics to high level programming.

Key Features:

  • Start by building an advanced robotic arm from scratch
  • Learn basics of AI and Computer Vision
  • Learn Arduino and Raspberry Pi

An Intelligent Robot You Can Compete With

Use the included computer vision kit to write and research your own AI programs, including face detection and recognition, face following and object tracking and pick-up. Or, try to recreate the included Tic-Tac-Toe playing code. When a game is over, it even resets the board so you don’t have to do it!

But Also Have Fun With

Program a song once, then play it again using EvoArm. Use our example provided note playing sequences, and stitch them together to play your own desired song. Or, write your own AI algorithm to come up with new songs and have EvoArm play them!


You can buy any typical robot arm kit off the internet. But once you build it, then what? Although you will find joy in constructing the EvoArm from scratch, the biggest delight is in the included programming and control software.

Once built and configured, EvoArm can be controlled and programmed from a phone or computer using an intuitively designed interface without requiring coding. For truly unlocking its features, it can be programmed with Python, C++ or any language that supports publishing ROS messages, or making HTTP requests.

What You Need to Get Started

Skills You Will Master

What's Included

Arduino Uno

Rasberry Pi
Raspberry Pi Zero W (with WiFi and Bluetooth capacity)
16GB Micro SD Card with pre-configured Raspbian Buster OS
Robot operating system (ROS)
OpenCV 3.4

Computer Vision Kit
Raspberry Pi camera
Camera mounting bracket and hardware
Up to 30fps streaming service

Impeccable Hardware
High power servo motors
LEDs, resistors, switch
45W wall power supply
Specially sourced screwdriver
Over 20 construction parts for Robot Arm
Cables and wiring

Included Accessory Kits and EvoApps
Towers of Hanoi kit
Whiteboard drawing kit
Warehouse Box stacking kit
Optional: AI-Enabled Tic-Tac-Toe kit 

EvoArm Specifications

Axes: 5 Degrees of Freedom + Gripper

Max Vertical Reach: 18″ or 460mm

Max Horizontal Reach: 14.5″ or 368mm

Payload: 100g at maximum extension

Power Supply: 45W Wall Adapter 110V

Repeatability: 1mm

Weight: 1.5lbs or 680g


A unique approach to building “real” physical apps. Design and print functional hardware, and program EvoArm to operate it. Your purchase includes the 4 EvoApps below.


Warehouse Box Stacking


Practice your coding skills


A fresh take on this classic computer science problem


Program the Tic-Tac-Toe Algorithm and play against EvoArm 

Building Instructions

This project comes with 16 detailed step by step tutorial videos of a few minutes each, which you can easily watch at your own pace. Whether you are already a pro, or a complete novice, *you* can build EvoArm yourself.

Step By Step Video Tutorials

1. Battery Pack Assembly
2. How to connect a Servo Extension Cable
3. Baseplate Assembly
4. Motor A Assembly
5. Motor B Assembly
6. Motor C Assembly
7. Motor D Assembly
8. Motor E Assembly

9. Motor F Assembly
10. Gripper Assembly
11. Arm Assembly
12. Arm Wiring Part 1
3. Arm Wiring Part 2
14. Raspberry Pi And Camera Assembly
15. Powering Up
16. Setup


Everything Included



Frequently Asked Questions

A Mac, Windows, or Linux laptop with administrative rights to install Arduino, Python and other programming tools. If you have a newer Mac or Windows laptop with only USB-C ports, you will need a USB-C to USB adapter to be able to connect the Arduino to your computer. Chromebooks will not work for our program.

We’re here to help! Submit a ticket, or ask a question in the community forum to get help fast.


All experience levels are welcome with us! Whether you’re someone with 0 experience, or someone who’s had their fair share of time with robotics, we’re confident you will learn something exciting with us.


Besides a laptop or computer:

  • Internet access to download required tools and utilities

After successfully completing the Robot Dog project, you will receive a downloadable PreDegree certificate you can use on your resume or LinkedIn to show your skills and knowledge in robotics.


Yes, try our Quadruped Robot Dog project.

All of our robots are manufactured in Mountain View, California, and designed in tandem with the curriculum for the best learning experience possible for hands-on robotics. 

We will do everything to make sure you have a flawless experience. 100% will be refunded if the kit has not been dispatched, or is returned without being opened. 

All other sales are final.

Yes. EvoArm comes with pre-installed advanced kinematics firmware. You can command individual joints or any combination simultaneously. Or you can use the available inverse kinematics function to command the destination position and orientation of gripper, and EvoArm will figure out and execute a smooth path.

Yes, EvoArm comes ready-to-go with a Raspberry Pi that is pre-programmed to control the arm hardware!

Yes! You can login to the included raspberry-pi and use the included python examples as a starting point. If you know ROS, you can make the raspberry-pi the ROS master and publish commands from your own computer. If you don’t know ROS, you can also send arm commands over simple an extremely simple HTTP Rest interface.

Yes, EvoArm is designed to be programmable with several different types of interfaces your custom app can use to control the arm, including ROS, and HTTP.

The payload limit is 100g at full extension.

Approximately 1mm 

Yes! The beauty of EvoArm lies not just in the hardware, but in the comprehensive, tested and well designed software that allows you to program it.

Yes. See videos of the arm solving Towers of Hanoi, or playing real Tic Tac Toe against a human player.

Yes, once setup and configured, EvoArm connects to your local wifi network so you can control and program it from your phone or computer.

Yes! Watch videos of the arm playing a song on a piano.

EvoArm is designed for educational and learning purposes, it is not suitable for an industrial setup.

Yes! One arm can control another over HTTP, or by publishing ROS messages to other arms. A program on an external computer could control multiple arms together with arbitrary complexity.

Yes! You can access the camera feed at upto 30 frames per second over HTTP for your own computer vision applications.

Yes, most certainly. You can access the camera frames, the arm’s current position information and send commands to move the arm, using C++ or python, or any language that supports making either simple HTTP calls or publishing ROS messages.

It comes with a special high-wattage wall power supply so you can keep EvoArm running. But if you want to make it mobile, it will also run on AA batteries and a portable USB phone charging bank.



-Arnar H.

We have been having trouble free use of the robot. I wanted to be sure you knew we are still enjoying the robot arm kit.

My youngest (the 11 year old) used it a bit to do the things that are easy (set up sequences, do exercises). I then gave it to my oldest (16 years old).  He’s a bit of a linux nerd and started looking into how the thing is put together (the rospi pubsub, the command stream sent to from the pi to the arduino etc).


-Terrazas A.

I can’t believe I didn’t think of that! Thanks. I was able to install XQuartz on my Mac and then succeeded to display a camera image.


-Kaleb N.

From the support and discussion forum:

What are the numerical ranges for the forward kinematic equation?

a: 10, 170 #base
b: 10, 170 #shoulder
c: 10, 148 #elbow
d: 10, 170 #wrist rotator
e: 10, 170 #wrist up/down
f:  40, 120 #gripper, smaller value closes

Thanks so much for the fast reply!

Customers Also Viewed