Software Setup

A few initializations should be made on the Pies before the software can be deployed on them. 0. If you haven't done that yet, enable your camera module and SSH from sudo raspi-config

1. Makes sure you have python3.7 as your default python3 in your Pi

python3 --version

1.b If you don't have python2 installed then you can use python and pip instead of python3 and pip3 from here on.

2. Install pipenv and tox globally using pip3 (use sudo to install them globally)

sudo pip3 install pipenv tox

3. To be able to install openCV via pip we need to install some additional dependencies globally as

sudo apt-get update
sudo apt-get install libcblas-dev libhdf5-dev libhdf5-serial-dev libatlas-base-dev libjasper-dev libqtgui4 libqt4-test

3.b We also need some other packages to dtach sessions from terminals and to keep our clock healthy even after long shutdown periods.

sudo apt-get install dtach htpdate

4. Clone this repository on the Pi

 git clone <repo https/ssh>

5. Install InfluxDB and Grafana according to this page then come back and continue from 6.

6. Install Aseba and Thymio firmware according to this page then come back and continue from 7.

7. Install CNN-based vision related software according to guide on the right.

8. Install and initialize Google APIs with guide on the left menu bar.

9. Cd in the clone and create an empty virtual environment and open the virtualenv

pipenv --python 3.7
pipenv shell

10. Install dependencies in the virtualenv after inside and in the same folder as setup.py. This step should be always repeated when the dependencies change during the development.

pip install -e .

the flag -e will chain your original .py files in the package install, so that you do not have to install the package every time you change something in the python code. After a git pull the changes will be applied to your code, and you can run the desired entrypoint right afterwards.