Thymio Robot Documentation: A total beginners guide to the galaxy

This manual explains how to use the code on this git when working with our equipment in the Robotlab of the VU in Amsterdam. We are currently working on a more general manual and explanation of the equipment you need to start your own experiments in your own lab. Many students use this code and develop it further. We try to have the most recent version online but it might happen that some explanations are double or not consistent. For questions you are welcome to contact j.v.heinerman AT vu.nl

Setting Up

Make sure you have a UNIX environment terminal to execute the commands needed to communicate with the Raspberry PI, e.g. Mac OSX or Linux and then connect to the WiFi.

• Channel: ThymioNet
• Password: 172luckytulip75B

Download the appropriate packages to execute and test the robots use

git clone https://github.com/ci-group/NEATThymio.git

Logging In

Check the IP address on the back of the robot you want to use, e.g. 192.168.1.72. Connect to the robot with SSH

ssh pi@<IP_addres>

You will be prompted for a password to connect ot Thymio

• Password: raspberry

NOTE: To avoid having to put in your password everytime share your public SSH key with the PI

ssh-keygen -t rsa
ssh-copy-id pi@<IP_addres>

[1] See: Passwordless SSH access

Executing Scripts

External execution

Pick your local booting script, e.g. start_one.sh and execute the script, e.g. foraging.py.

./start_one.sh foraging.py 192.168.1.72 1

Local multiple execution

Pick your local booting script, e.g. start_all.sh and open bots.txt

cd NEATThymio
cd src
vim bots.txt

Type in all IPs of PIs you want to start and make sure there is an empty line in the bottom, e.g:

192.168.1.52
192.168.1.72
192.168.1.62

Execute the script you want, e.g. foraging.py

./start_all.sh foraging.py

Synchronize

Use rsync to post on the PI and pick a robot, e.g. 192.168.1.52 Pick a file on your computer, e.g. test.txt and pwd the path. Rsync it with the PI

rsync <path_to>/test.txt pi@192.168.1.72:~/

Use rsync to pull from the PI

rsync pi@<IP_address>:~/test.txt <path_to_copy>

Clean Install

Get an SD-card, preferably 8GB or more and format it to FAT32. Install Raspbian image Find the correct IP address of the PI

nmap -sn 192.168.1.*

Install the needed packages on PI

 sudo apt-get update
 sudo apt-get install rsync python-opencv python-numpy python-dbus python-gobject python-picamera dbus dbus-x11
 sudo apt-get install libboost-dev libqt4-dev qt4-dev-tools libqwt5-qt4-dev libudev-dev cmake g++ git make

NOTE: If the last one does not work try the following commands followed by the above install again

 sudo apt-get clean
 sudo apt-get -f install

Download the Aseba images and rsync the .deb file to the PI. Run the following command:

sudo dpkg -i aseba_1.4.0_armhf.deb

Setup PI Wifi

Stick a ETHERNET cable in your PI and find the correct IP address of the PI

nmap -sn 192.168.1.*

SSH with your PI and:

vim /etc/network/interfaces

Replace file content with the following

auto wlan0

auto lo

iface lo inet loopback
iface eth0 inet dhcp

allow-hotplug wlan0
iface wlan0 inet static
address 192.168.1.XX
netmask 255.255.255.0
gateway 192.168.1.1
dns-nameservers 8.8.8.8
wpa-conf /etc/wpa_supplicant/wpa_supplicant.conf
iface default inet dhcp

NOTE: Replace the XX with your IP address

Also alter this file:

vim /etc/wpa_supplicant/wpa_supplicant.conf

and replace file content with the following:

ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
update_config=1

network={
		ssid="ThymioNet"
		psk="172luckytulip75B"
		proto=RSN
		key_mgmt=WPA-PSK
		pairwise=CCMP
		auth_alg=OPEN
}

Using the logs

Go the log directory of your NEATThymio map and change bots.txt to include the IP(s) of the PI you want the logs off.

Run get_logs.sh locally

./get_logs.sh bots.txt

You will get an error that a map does not exist. Then create that map locally and repeat previous step again:

./get_logs.sh bots.txt

Either create CSV files or visual representations of each NN with:

1. Get the .json file you want

2. Put it in the same directory as export_data.py or neural_net_vis.py

   python export_data FILE.json

Authors

• Tjarco Kerssens
• Cees Schouten