Skip to content

Latest commit

 

History

History
157 lines (137 loc) · 8.49 KB

README.md

File metadata and controls

157 lines (137 loc) · 8.49 KB
Raw

How to cite this software

The 2D quadrotor landing simulation is implemented in JavaScript and can be added to any website. In this repo, the website is served using a Python Flask application. Various versions of the simulation have been used in the following projects:

  • Emily Jensen, Sriram Sankaranarayanan, and Bradley Hayes. 2024. Automated Assessment and Adaptive Multimodal Formative Feedback Improves Psychomotor Skills Training Outcomes in Quadrotor Teleoperation. International Conference on Human-Agent Interaction (HAI '24). https://doi.org/10.1145/3687272.3688322
  • Madeleine S. Yuh, Ethan Rabb, Adam Thorpe, and Neera Jain. 2024. Using Reward Shaping to Train Cognitive–based Control Policies for Intelligent Tutoring Systems. 2024 American Control Conference (ACC). link to paper
  • Sooyung Byeon, Dawei Sun, and Inseok Hwang. 2021. Skill-level-based Hybrid Shared Control for Human-Automation Systems. 2021 IEEE International Conference on Systems, Man, and Cybernetics (SMC). https://doi.org/10.1109/SMC52423.2021.9658994

If you use this software in your own project, please use the following citation:

@misc{quadrotor-simulation,
  author = {Byeon, Sooyung and Jensen, Emily and Yuh, Madeleine S. and Sankaranarayanan, Sriram and Jain, Neera and Hwang, Inseok},
  title = {Quadrotor Landing Simulation},
  year  = {2024},
  url   = {https://github.com/cairo-robotics/llm-multimodal-drone-feedback},
}

Using this repo

This README outlines the steps to deploy a Flask application on a Red Hat Linux instance using Gunicorn on port 5000 and Nginx as a reverse proxy. Follow these instructions to set up your project environment, deploy your Flask application, ensure it runs successfully with Gunicorn, and configure Nginx.

Warning: These steps make it look a lot easier than I experienced trying to get this set up the first time. I tried to distill my multiple false starts into a coherent document. You may need to try rearranging some steps or reloading the firewall/Nginx multiple times to get things working. Hopefully your journey will be easier than mine!

Preliminary Steps

  1. Request a managed instance from the CS IT team. Make sure you ask them to make a SSL certificate and key for you. The instance should be public (not internal). For the rest of the steps, I will assume the instance is called dronefeedback.colorado.edu.
  2. While on the UCB VPN, log into your instance. For the rest of the steps, I will assume the identikey is emje6419.
ssh emje6419@dronefeedback.colorado.edu

System Update and Package Installation

Update system packages and get python set up:

sudo yum update -y
sudo yum install git python3-devel -y
sudo yum groupinstall 'Development Tools' -y
pip3 install virtualenv

Configure Github CLI

  1. Install the Github CLI package:
sudo dnf install 'dnf-command(config-manager)'
sudo dnf config-manager --add-repo https://cli.github.com/packages/rpm/gh-cli.repo
sudo dnf install gh
sudo dnf update gh
  1. Create a (Classic) personal access token. Make sure you have enabled 'repo', 'read:org', and 'workflow' scopes.
  2. Log in to GitHub using personal access token. When prompted, choose "github.com", "HTTPS", and authenticate using authentication token.
gh auth login

Set Up Flask Project

  1. Set up your virtual environment in your project directory
mkdir myflaskapp
cd myflaskapp
python3 -m virtualenv venv
source venv/bin/activate
  1. Clone this repository and install dependencies
git clone https://github.com/cairo-robotics/llm-multimodal-drone-feedback.git
cd llm-multimodal-drone-feedback
pip install -r requirements.txt
  1. Set up your OpenAI key (replace XXX with your key)
echo OPENAI_API_KEY="XXX" >> .env
  1. Open relevant ports and reload firewall
sudo firewall-cmd --permanent --add-port=80/tcp
sudo firewall-cmd --permanent --add-port=443/tcp
sudo firewall-cmd --zone=public --add-port=5000/tcp --permanent
sudo firewall-cmd --reload

  1. Get your server's IP address using curl -4 icanhazip.com

  2. Connect Flask and Gunicorn. You will know the server is running if the command line blinks like it is busy and you can't enter other commands. If nothing seems to happen or other issues come up, you can check the log file you specified for details.

export FLASK_ENV=development
export FLASK_APP=server.py
gunicorn --workers 3 --bind 0.0.0.0:5000 wsgi:app --log-file /home/emje6419/myflaskapp/mygunicorn.log

  1. Check that you can access your project at http://<server-ip>:5000.

  2. Shut down server for now using CTRL+C.

Generate SSL Certificate

We have had inconsistent luck getting CS IT to generate an SSL certificate. If they direct you to use campus OIT, then follow these steps.

  1. Request an SSL certificate through OIT's website.
  2. You will get an email that instructs you to go to Sectigo. It will have other steps as well. You'll fill out a form that asks for a CSR (Certificate Signing Request). This is a file that has information about your system. If it doesn't exist on your system, you will have to generate one.
  3. To find out if there is one already, go to your root directory and enter: sudo find / -type f -name "*.csr"
  4. If there is no result, you have to create one. To make one, first generate a private key on the remote server:
sudo yum install openssl
openssl genrsa -out mydomain.key 2048
  1. Generate the CSR on the remote server: openssl req -new -key mydomain.key -out mydomain.csr. It will ask for your Country Name (2 letter code) which here is US. When you generate the challenge password, make sure you make note of it! you will need it later when you get your ssl certificate.
  2. To securely copy the csr to your local machine so that you can upload it to Sectigo, locate the csr on your host: sudo find / -type f -name "*.csr". This will give the path for the file.
  3. Then open a local terminal window (try using Git Bash) and copy the file.
scp emje6419@dronefeedback.colorado.edu:<path for csr file.csr> 'C:/Users...<local destination folder>'
  1. You can then upload the CSR to the Sectigo form.

Set up Nginx

Instructions are based off of this website.

  1. Restart your flask app
source venv/bin/activate
gunicorn --workers 3 --bind 0.0.0.0:5000 wsgi:app --log-file /home/emje6419/myflaskapp/mygunicorn.log
  1. Install Nginx package
sudo yum install nginx
  1. Move the configuration file from this repo to the Nginx directory. Make sure to update server_name with your specific name and the SSL certificate/key locations.
sudo mv myflaskapp.conf /etc/nginx/conf.d
  1. Test and Start Nginx
sudo systemctl enable nginx
sudo systemctl start nginx
  1. Check that you can see the system test page by navigating to http://<server-ip>.

Final Steps

  1. Start a tmux session
sudo yum install tmux -y
tmux new -s mysession
  1. Make sure the experiment is set to the appropriate feedback condition in the Experiment.js file.
  2. Start your server in the tmux session (your virtual environment should already be activated)
gunicorn --workers 3 --bind 0.0.0.0:5000 wsgi:app --log-file /home/emje6419/myflaskapp/mygunicorn.log

If you are feeling really fancy, you can try to create a service that will automatically run this. I gave up after struggling with SELinux policy permissions for several hours.

  1. Detatch from your tmux session using CTRL+B then d. You can attach later using tmux attach -t mysession.
  2. Check you can access your project at https://dronefeedback.colorado.edu. Congrats, you did it!

Transfer Data Back to Windows

  1. Open Git Bash
  2. Make a folder on Windows where you want your data to be stored
  3. Use SCP to copy files
scp -r emje6419@dronefeedback.colorado.edu:/home/emje6419/myflaskapp/llm-multimodal-drone-feedback/static/data 'C:/Users/Emily Jensen/OneDrive - UCB-O365/Drone Feedback Data'