Sleep-Focused Hospital Scheduler

Problem

This project is designed to help patients at capital region hospitals such as Troy's Samaritan Hospital. When nurses do rounds, they often wake up patients. This is especially true at night when patients are trying to sleep. While nurses must check on patients, they should do so in a way that minimizes the amount of sleep patients lose.

Solution

This project schedules the nurses in a hospital. Its goal is to maximize pationt sleep, and thus improve patient outcomes. The scheduler's main advantage over a traditional rounds system is that it tries to align nurse schedules. Rather than a patient being woken up twice in a normal system, under a schedule computed by this system the patient would be woken up only once because both nurses enter the room at the same time.

This project includes a backend and a frontend. The backend is written in Python and uses Google's OR-Tools to solve the scheduling problem. The frontend is a windows application written in Flutter and provides a user interface for inputting tasks into the scheduler and viewing the solution.

How to run

Requirements

• Python >=3.13
• Flutter >=3.38
• Visual Studio 2022 with Desktop development with C++ workload installed

Backend

Create venv and then install the dependencies with

pip install -r requirements.txt

Run the server

flask run

Frontend

Get dependencies

flutter pub get

Run the frontend

flutter run

File structure

backend/          # Python backend
  app.py          # Main Flask application
  requirements.txt  # Python dependencies
hospital_scheduler_frontend/         # Flutter frontend
  lib/            # Flutter application code
  pubspec.yaml    # Flutter dependencies

Please note: while it is possible to run the backend on a separate machine from the frontend, the system is configured for running on a single machine. The backend is heavily optimized so it can run on a nonpowerful laptop. The original algorithm I implemented on the morning of the hackathon used 12 GB of memory but I optimized it to use less than ~50 MB of memory.

Sources

This paper indicates that, "Poor sleep is a pervasive problem for hospitalized patients and can contribute to adverse health outcomes." Herscher, M., Mikhaylov, D., Barazani, S., Sastow, D., Yeo, I., Dunn, A. S., & Cho, H. J. (2021). A sleep hygiene intervention to improve sleep quality for hospitalized patients. The Joint Commission Journal on Quality and Patient Safety, 47(6), 343-346.