MiniBOT candy picking demo

This demo presents how to use NexCOBOT MiniBOT to pick candy with AWS SageMaker as inference engine.

In this demo, we use AWS SageMaker to label our candy dataset, train a SSD model, and setup endpoint to do inference. I will create another repo for the detail.

Here is the result of the inference (also a youtube link of demo video):

System Architecture

MiniBOT software environment setup for candy picking

The OS platform of MiniBOT controlller is Windows 7 32-bit

Pre-requirement

• Install NexMotion Studio to get NexMotion C/C++ API and NexMotion.dll.
• Create a project to setup MiniBOT Configuration from NexMotion Studio.
• For operating real robot arm, do homing process in NexMotion Studio first.

Clone Nexmotion Python API and this demo jupyter notebook.

In windows system, you can install Git to run following command in Git Bash

# Go to Documents folder
cd Documents
# clone NexMotion Python API
git clone https://github.com/RobinCPC/NexMotion-Python_API.git
cd NexMotion-Python_API
# cloen this demo repo
https://github.com/RobinCPC/minibot_candy_pick_demo.git

AWS Setup

• Install AWS CLI version 1 (2 only for 64-bit)
  https://docs.aws.amazon.com/cli/latest/userguide/install-windows.html#install-msi-on-windows
• Set up aws configuration https://docs.aws.amazon.com/cli/latest/userguide/cli-configure-files.html
  For configure, open Windows command prompt (CMD), and enter aws configure. Then, will show parameter for setup.

REM ntoe: you should got your own key from your AWS account
aws_access_key_id=XXXAIOSFODNN7EXAMPLE
aws_secret_access_key=wJalrXUtnFXXX/K7MDENG/bPxRfiCYEXAMPLEKEY
region=us-west-2
output=json

Python Setup

• Install python 2.7 from python.org or use conda
  https://www.python.org/downloads/release/python-2717/

• Setup virtual environment (In MiniBOT controller)
  Open Windows CMD, then

REM install virtualenv
pip isntall virtualenv
REM cd to Document or where you store your code
cd \Users\<your_name>\Documents\
REM create folder 
mkdir python-virtual-env
cd python-virtual-env
REM create virtual env
virtualenv minibot-env

• Activate environment and install necessary package for candy picking project
  Still in Windows CMD

REM activate python virtual env
cd Documents\python-virtual-env
minibot-env\Scripts\activate
pip install -r requirements.txt

• Back to this repo folder and launch jupyter notebook

cd Document\NexMotion-Python_API\minibot_candy_pick_demo
jupyter notebook

In addition, I also use ipywidgets to create robot controller WebUI in Jupyterlab.
Here is the screenshot (also a youtube link of demo video):