Open Source Ventilator Remote Monitor Project - Rails App

This git repo is for the code on the Heroku Cloud Server & Web Browser

Architecture Overview

Heroku Cloud Server & Web Browser Repo	Raspberry Pi (Network Adapter) Repo	Ventilator (Arduino) Repo
Ruby on Rails	Flask	Arduino Board
Javascript	Python	C++

Goal

Our goal is to quickly develop a remote monitoring interface for low-cost rapidly-manufactured ventilators currently being developed to provide emergency relief in the Covid-19 Pandemic. We strive for a lightweight, interoperable, and reliable interface. Currently the interface provides ventilator remote monitoring only (no control).

This software is not approved to be used in a medical setting.

Live Demo

Check out the live demo/concept.

How do I set up my computer for development?

1. Set up your computer to run a Ruby on Rails develpment environment. This is quite an involved process - thankfully you only need to do it once in a while. My preferred method is to use the thoughtbot laptop script for mac. Gorails also has a nice setup guide for multiple operating systems. Be sure to install Rails 6.0.2.1.
2. Install yarn: brew install yarn
3. Install webpacker: rails webpacker:install
4. Install postgres: brew install postgres
5. Clone this repo: git clone https://github.com/Open-Ventilator-Remote-Monitoring/ventilator-remote-monitoring.git
6. cd ventilator-remote-monitoring
7. Create the rails database: rails db:create
8. Run bundler: bundle install
9. Migrate the database: rails db:migrate
10. Start a rails development server: rails s
11. You should now be able to navigate to http://localhost:3000 and view the live rails site

Having problems? Post on the Slack Channel and we'll help you out

Creating the first Administrator

If this is your first use since the User database table was created, you will need to set up an Administrator. Here are the steps:

1. Start the web server and open a browser to http://localhost:3000

2. In the top-right of the page, click on 'Sign in'

3. On the Log In screen, click on 'Sign Up'. Type in your email address, name and the password you want to use (twice) and click 'sign up'.

4. In the directory where the project is located, run the rails console-

   rails c

5. Execute the following:

   me = User.find(1)

   Ensure the record displayed includes your email address. If not, and you are not sure of how to find your record, stop here and ask for help.

6. Execute the following:

   me.admin!

   This should change your roll to admin and update the database.

7. Type

   exit

   Now when you go back to the web page and log in you should be an administrator.

8. As soon as you have created your first organization, please navigate to http://localhost:3000/users, find your email address under the heading "Users awaiting Assignment to an Organization", and click the "assign" button. Select an Organization from the drop-down list and click 'update user'.

API

A call to /api/v1/ventalators returns a single organization, all of the clusters in that organization and all of the ventilators in those clusters. The organization used is the organization assigned to the person who is logged in. If the call is made from a browser which is not logged in, or there is no organization associated with the current user, then the following is returned:

{
  "error": "User is not logged in or is not associated with an Organization."
}

Note: This could change to a default 'demo' data to be used by a demo web page. But it may be better for the demo page just to provide some demo data locally.

The serialization library used serializes the data as a graph using the json:qpi spec.

Note: Beware when matching nodes, in that it is not enough to match on id, as the same id numbers are used in organization, clusters and ventilators. Match on both type and id.

Here is an example of the JSON returned:

{
  "data": {
    "id": "3",
    "type": "organization",
    "attributes": {
      "name": "Honnah Lee Medical Center"
    },
    "relationships": {
      "clusters": {
        "data": [
          {
            "id": "6",
            "type": "cluster"
          },
          {
            "id": "7",
            "type": "cluster"
          }
        ]
      }
    }
  },
  "included": [
    {
      "id": "6",
      "type": "cluster",
      "attributes": {
        "name": "East Wing"
      },
      "relationships": {
        "ventilators": {
          "data": [
            {
              "id": "15",
              "type": "ventilator"
            },
            {
              "id": "16",
              "type": "ventilator"
            },
            {
              "id": "17",
              "type": "ventilator"
            }
          ]
        }
      }
    },
    {
      "id": "7",
      "type": "cluster",
      "attributes": {
        "name": "West Wing"
      },
      "relationships": {
        "ventilators": {
          "data": [
            {
              "id": "18",
              "type": "ventilator"
            },
            {
              "id": "19",
              "type": "ventilator"
            }
          ]
        }
      }
    },
    {
      "id": "15",
      "type": "ventilator",
      "attributes": {
        "id": 15,
        "name": "Room E-1 Bed 1",
        "hostname": null
      }
    },
    {
      "id": "16",
      "type": "ventilator",
      "attributes": {
        "id": 16,
        "name": "Room E-2 Bed 1",
        "hostname": null
      }
    },
    {
      "id": "17",
      "type": "ventilator",
      "attributes": {
        "id": 17,
        "name": "Room E-3 Bed 1",
        "hostname": null
      }
    },
    {
      "id": "18",
      "type": "ventilator",
      "attributes": {
        "id": 18,
        "name": "Room W-1 Bed 1",
        "hostname": null
      }
    },
    {
      "id": "19",
      "type": "ventilator",
      "attributes": {
        "id": 19,
        "name": "Room W-1 Bed 1",
        "hostname": null
      }
    }
  ]
}

fast_jsonapi

This project uses fast_jsonapi from Netflix for JSON serialization in Ruby. Netflix is no longer maintaining the libray and the commuinity has taken over. That project is (here)[https://github.com/fast-jsonapi/fast_jsonapi]. If you encounter any issue, it may be best to move to that project.

Deserialization

On the clint, these few lines of code will deserialize the data above:

  import Jsona from 'jsona'

  const dataFormatter = new Jsona()
  const organization = dataFormatter.deserialize(response.parsedBody)

This results in something we can more easily work with (relationshipNames keys have been removed):

{
  "type": "organization",
  "id": "3",
  "name": "Honnah Lee Medical Center",
  "clusters": [
    {
      "type": "cluster",
      "id": "6",
      "name": "East Wing",
      "ventilators": [
        {
          "type": "ventilator",
          "id": 15,
          "name": "Room E-1 Bed 1",
          "hostname": null
        },
        {
          "type": "ventilator",
          "id": 16,
          "name": "Room E-2 Bed 1",
          "hostname": null
        },
        {
          "type": "ventilator",
          "id": 17,
          "name": "Room E-3 Bed 1",
          "hostname": null
        }
      ]
    },
    {
      "type": "cluster",
      "id": "7",
      "name": "West Wing",
      "ventilators": [
        {
          "type": "ventilator",
          "id": 18,
          "name": "Room W-1 Bed 1",
          "hostname": null
        },
        {
          "type": "ventilator",
          "id": 19,
          "name": "Room W-1 Bed 2",
          "hostname": null
        }
      ]
    }
  ]
}