Edog - Mini 3D printed Quadruped Robot ✨

Description

**** ᴛʜɪs ᴘʀᴏᴊᴇᴄᴛ ᴡɪʟʟ ʙᴇ ᴏᴘᴇɴsᴏᴜʀᴄᴇ ɪɴ ᴊᴜɴᴇ 𝟸𝟶𝟸𝟷 ****

🦄 My dream is to create a community of eDog makers 🌈

I wish that everyone could build their own robot around the world and share with the community !

To reach this goal, eDog will be opensource as soon the project is completed ! Everything you need will be documented and explained

Please share and subscribe to make it happen 🥳

Details

Challenges / Constraints :

100% 3D printable (without support)✔️
Under 300gr weigh✔️
Under 100$ budget ✔️
Easy to build ✔️
Linux embedded✔️
ROS2 Environment ⌛
Modular design / components ✔️
Starter kit sell ⌛

Achieved:

The robotic frame (3D printed, easy to assemble, workable and viable)
A stable electronic environement (5V powered, 8A max, Raspberry pi, 16 servo driver)
Inverse Kinematics
Socket service to control the robot from a network
Walking motion v.1

On going

Software Framework
Walking motion v.2

To do:

Explode data from the gyro module
Refactoring the embedded code
Multi edog control

For the eurobotCup 2021:

Obstacle detection (using a simple HC-SR04 Ultrasonic Distance Sensor)❌
Embedded webcam❌
Computer vision for object recognition and travel estimation❌
Arm actuator (using pump and suction cup) ❌

Components

1 × Raspberry pi Zero WH

1 × Driver Servo - PCA9685 16' canaux 12-bit pwm- I2C

1 × Power supply - HENGE 8A UBEC 5v / 6v / 7.4v 7V-25.5V Input for 2-6 Lipo

1 × GY-521 MPU6050 3 Axes Accelerometer Gyroscope

8 × Servo Motors - VOTIK 7455 MG-D 9g

Project Logs

Collapse

Beacon Controller
Gaultier Lecaillon • 02/14/2021 at 11:53 • 0 comments

As explain, edog will participate in the Eurobot cup.

To do so, I need to know the robot position in (x, y) on the table. But I can't use embedded lidar or using triangulation due to its small size and its limited capacity to carry heavy components.
So What I will attempt to do is using a beacon with a camera and estimate the position giving a video flux.
This beacon is composed with an NVIDIA Jetson Nano 2gb board, a TPLink Nano Router, a 7inch touchscreen and a DC converter to 5v and 5A. that allow to used 3s to 6s lipo to power up the sytem.

Here an exemple of the team ARIG Robotique (more infos here)

Step #1: original video
Step #2: Optical distortion usgin OpenCv in python
Step #3: projection in an orthonormal coordinate system
Exploded view of all edog components
Gaultier Lecaillon • 01/31/2021 at 00:00 • 0 comments

All the stl files will be drop after the French Eurobot cup, the 1st June 2021.

here you can see all the part involves, Everything has been design for an easy build and easy printed without support needed.

Only 4 * M3 screw are needed for the frame ! 🔩
It makes edog very simple to build, so everyone can build his own easely
eDog Walking demo 31/01/2021
Gaultier Lecaillon • 01/30/2021 at 23:40 • 0 comments

All legs are controlled by an inverse kinematics algorithm. ROS2 will be implement soon !