- Pre-installed Donkey Car Software
- Support Donkey Car mobile app
- Supports RPI4B
- Pre-installed tensorflow 2.5
- Pre-installed coral library
- Jupyter Lab
- Connected SSID, system password is displayed via Donkey HAT OLED
Supports Robohat MM1Built-in wifi hotspot
Date | Platform | Donkeycar version | Download |
---|---|---|---|
2023-11-14 | Raspberry Pi 4B / Bulleye | v5.0-dev3 | Download |
2023-04-06 | Raspberry Pi 4B / Buster | v4.3.22 | Download |
2022-11-03 | Raspberry Pi 4B / Buster | v4.3.22 | Download |
2022-10-27 | Raspberry Pi 4B / Buster | v4.3.22 | Download |
2022-06-02 | Raspberry Pi 4B / Buster | v4 | Download |
2021-12-07 | Raspberry Pi 4B / Buster | v4 | Download |
2021-07-12 | Jetson Nano | v4 | Download |
2021-05-27 | Raspberry Pi 4B / Buster | v4 | Download |
2021-05-26 | Raspberry Pi 4B / Buster | v4 | Download |
2021-05-20 | Raspberry Pi 4B / Buster | v4 | Download |
2021-02-04 | Raspberry Pi 4B / Buster | v4 | Download |
2020-10-14 | Raspberry Pi 4B / Buster | v3 | Download |
2020-07-18 | Raspberry Pi 4B / Buster | v3 | Download |
A random 6-digit password is generated during the first boot. This password is used to access
- Jupyter Lab
- Mobile App
- SSH
There are two ways to retreive this password:
- The password will be displayed through the OLED display of the Donkey HAT
- The password is stored in /home/pi/donkey.cfg
To change the password, do not directly edit /home/pi/donkey.cfg. Editing this file directly will result in permanent loss of the shell login if you have not written it down. Use /opt/donkeycar-images/changepw.sh instead. This script will change the password of the raspbian user, jupyter-lab and the mobile app password in one go.
- username: pi
- password:
- username: nano
- password: asdfasdf
http://donkey-xxxxxx.local:8888
sudo git clone https://github.com/sctse999/donkeycar-images /opt/donkeycar-images
sudo ln -s donkey-init.service /etc/systemd/system/donkey-init.service
pip install jupyterlab
sudo systemctl enable jupyter-lab.service
- Enable SSH
- Enable Camera
- Enable I2C
/etc/systemd/system/jupyter-lab.service
[Unit]
Description = JupyterLab
[Service]
PIDFile = /run/jupyter.pid
ExecStart = /home/pi/env/bin/jupyter-lab
User = pi
Group = pi
WorkingDirectory = /home/pi
Restart=always
RestartSec=10
[Install]
WantedBy = multi-user.target
The reset script will reset the current system to a clean state.
# Run the following command
/opt/donkey-images/pre-release-v4.sh
-
Hostname should be renamed to donkey-xxxxxx where xxxxxx is the last 6 character of the wlan0 hardware address Use
ifconfig
to check wlan0 hardware address -
2.4 or 5ghz band should be configurable using/boot/donkey.cfg
-
Jupyter lab should be accessible via
http://<ip>:8888/
-
Donkey car console should be running. Check
http://<ip>:8000/vehicle/status
-
Mobile app should be able to finish one whole cycle of setup
-
Calibration should work
-
Drive UI should work
-
Training should work
origin: https://github.com/robocarstore/donkeycar
v4 branch: robocarstore_v4
v3 branch: robocar_dev
origin: https://github.com/robocarstore/donkeycar-console
branch: dev
If you intend to use this project for making money, you must obtain our consent before you do so. Contact us at sales@robocarstore.com
- Flip camera by default
- Fix recording not working issue by updating to the latest.
- Recreate donkeycar app for calibration and copilot support
- Fix gunicorn.service symlink location
- numpy upgrade to 1.20.3 fixing
- Fix donkey makemovie
- Random 6-digit system password
- System password is displayed via Donkey HAT OLED display
- Jupyter lab is password protected
- Installed with donkey car v4.3.22
- Disable donkey car LED part
- Support Donkey HAT
- Added
oled-hostname-ip.py
- Enable OLED part by default
- Switched to TF2.5
- Removed low battery auto shutdown cron job
- Installed keras-vis, moviepy and scikit-image to support tub movie generation
- pre-installed edge tpu dependencies
- Tuned pigpio to 10microsecond sampling rate (https://forum.radxa.com/t/pigpiod-eating-7-cpu-nonstop/3911)
- Installed donkey gym by default
- Installed Tensorflow 2.5
- Upgraded to donkey 4.3
- Updated /opt/donkeycar-console to the newest version (commit 01adc2ca841f5d1c17332d22e02bf260dbb1d9f4)
- image version now written to /boot/version
- add two router config to wpa_supplicant.test.conf
- default hotspot band to 5Ghz
- Added ping vehicle status to avoid slow first time loading of vehicle status page
- Added /boot/donkey.cfg to allow Windows based machine to edit this file to update configuration
- Fixed Jupyter Lab missing installation
- Added hotspot band parameter in
donkey.cfg
- Automatic rename ssid based on hardware address on every reboot
- Update donkey car to v4.2.0
- Compatible with Donkey Car Controller 1.3.2
- Single python venv at /home/pi/env. Both donkey console and donkey car software use this venv.
- Update donkey car to v3.1.5
- Update to latest donkeycar-console
- Fine tune myconfig.py value
- Install Google Coral Edgetpu library
- Installed jupyter lab as a service
- Installed ffmpeg and keras-vis
apt-get install ffmpeg
pip install git+https://github.com/sctse999/keras-vis.git
- Installed donkey console v2 under ~/
- Installed donkey console as systemd service
- Rename init script to donkey-init.sh
- Added ~/donkey.cfg for config reset
- donkey-init.sh will rename hostname when DONKEY_RESET=True in donkey.cfg
- donkey-init.sh will reset wpa_supplicant.conf
- Change raspap country to US and use 80211.ac 5Ghz
- Install switch-network.sh to crontab
- Performed OS upgrade
- Removed vim
- Added librealsense support (build 2.33.1)
- Enabled serial port for MM1 and installed pyserial
- Upgraded pip to 20.0.2
- Installed vim
- Adopted https://github.com/amix/vimrc