Back to Beril's profile
16views

Operation Ward Zero – Nursio SmartCare AI V3

A smart home-care monitoring prototype that simulates patient vitals, explains health-risk levels and helps caregivers respond through an integrated dashboard.

Beril Godjerin
  • Video of it working
  • Problem, approach and result explained
  • Public source code
  • Built during the challenge
  • Answered the full brief· 6 of 6
  • Expert reviewed
Operation Ward Zero – Nursio SmartCare AI V3

3

Simulated patient scenarios

4

Vital signs monitored

3

Risk levels with explanations

Walkthrough

Overview

The problem

Patients receiving care at home may not have access to continuous monitoring and timely caregiver attention like they would in a hospital. Caregivers may find it difficult to notice changes in vital signs, track patient conditions and respond quickly when readings become concerning.

Operation Ward Zero addresses this challenge by demonstrating how a bedroom can function like a connected home-care monitoring environment. The goal is to bring simulated patient monitoring, understandable risk alerts and caregiver response into one simple platform.

The approach

I developed Operation Ward Zero V3 as a full-stack home-care monitoring prototype using HTML, CSS, JavaScript and Node.js.

The application simulates patient sensors and generates heart rate, SpO₂, body temperature and systolic blood pressure readings every three seconds. The readings are sent to a Node.js backend and displayed on a monitoring dashboard.

A transparent rule-based risk engine evaluates the readings and assigns LOW, MEDIUM or HIGH risk levels with explanations. The platform includes vital trend charts, caregiver alerts, alert acknowledgement and resolution, caregiver check-in/check-out, and multilingual caregiver notes in English, Tamil and Hindi.

I also implemented JSON-based persistence and a limited browser-side offline queue that retries synthetic reading submissions after connectivity is restored.

The complete workflow is Sense → Understand → Alert → Respond.

The result

The result is a working browser-based home-care monitoring prototype with simulated patient readings, a Node.js backend and a caregiver dashboard.

The system demonstrates three patient scenarios: Stable Patient, Gradual Deterioration and Critical Readings. It displays vital signs, explains LOW/MEDIUM/HIGH risk assessments, and allows caregivers to acknowledge and resolve alerts.

The offline recovery test demonstrated that 74 queued synthetic reading submissions could be retried after reconnection, with the pending queue returning to zero and the backend accepting the submissions.

The project includes technical documentation, setup instructions, a public GitHub repository, application screenshots and a three-minute demonstration video.

All patient readings are synthetic. The risk engine is rule-based and is not a trained or clinically validated machine-learning model.

Reflection

Building Operation Ward Zero helped me understand how IoT simulation, backend services, frontend dashboards and caregiver workflows can work together in a practical home-care scenario.

One important challenge was maintaining the monitoring workflow when connectivity was interrupted. I implemented a browser-side queue and tested its recovery after reconnection.

I also learned the importance of explainability and clinical safety. The current prototype uses predefined rules rather than trained clinical AI, and caregiver notes are not yet included in risk calculations.

Future improvements could include secure caregiver authentication, stronger offline and power-loss resilience, integration with real medical sensors, multilingual note analysis and carefully validated predictive models.

This project is an educational prototype, not a medical diagnostic or treatment system.

AI tools used

ChatGPT

Used ChatGPT for project planning, application development assistance, Node.js debugging, documentation, testing guidance, and preparation of the project demonstration.

Links & files

Artifacts

1

Gallery

5