ROMAP

Team Name

ROMAP Team

Timeline

Summer 2025 – Fall 2025

Students

• Zaineel Mithani – Computer Science (Honors)
• Wrewoina Belai – Computer Science
• Barbara Green – Computer Science
• Richmond Onyeagba – Computer Science
• Pritesh Parekh – Computer Science
• David Saucedo – Computer Science

Sponsor

Arlington Sunrise Rotary Club

Abstract

The Rotary Operations Management & Automation Platform (ROMAP) is a comprehensive software system designed to modernize attendance tracking and data submission for Rotary Clubs. The project introduces a Bluetooth Low Energy (BLE) proximity-based attendance system, enabling hands-free, automatic member check-ins with 97% accuracy. Furthermore, ROMAP utilizes a novel GPT-4 Vision automation system to bypass the lack of an official API, intelligently populating web forms on the district portal with a 95% success rate. The system includes a mobile application, a Node.js backend, and a PostgreSQL database, effectively eliminating manual data entry and significantly reducing administrative overhead.

Background

Rotary International clubs are required to meticulously track member attendance and volunteer service hours for district-level reporting. However, the current process is inefficient, relying on manual sign-in sheets and self-reporting, followed by manual data entry into an antiquated district web application. This district platform lacks an Application Programming Interface (API), creating a bottleneck that forces club officers to spend hours transcribing data, a process highly susceptible to human error. ROMAP addresses this by engineering a modern solution that automates data collection via mobile devices and utilizes AI to interface with the legacy web portal.

Project Requirements

• Cross-Platform Mobile Application: A unified app for iOS and Android serving as the primary interface for members.
• BLE Proximity Check-in: An automated, hands-free check-in system using Bluetooth Low Energy beacons.
• QR Code & Manual Check-in: Alternative check-in methods generating unique, secure QR codes for scanning.
• AI-Driven Automation: A module using GPT-4 Vision to visually interpret and populate district web forms without an API.
• Volunteer Hour Logging: An in-app form for members to self-report service hours and activities.
• Secure Backend Infrastructure: A Node.js API and PostgreSQL database to serve as the central repository for club data.
• Role-Based Access Control: Distinct permissions for Administrators, Officers, and regular Members.
• Data Dashboards: Visual interfaces for viewing attendance records and volunteer summaries.
• Offline Capability: Ability to cache check-ins locally and sync when connectivity is restored.
• Error Handling & Notifications: A system to handle automation failures with exponential backoff and notify admins

Design Constraints

The system must interact with a district web portal that has no API, dynamic element IDs, and
frequent UI changes, constraining the automation approach to visual AI analysis rather than
traditional scraping

• Privacy & Legal Considerations: The handling of Personally Identifiable Information (PII) and location data requires strict adherence to data privacy standards, including informed consent and data minimization.
• Constructability/Hardware Costs: The implementation of the proximity system is constrained by the cost of BLE beacons and the need for specific hardware configurations (UUID, transmission power).
• Functionality (Mobile Restrictions): The mobile application is constrained by operating system limitations, specifically Apple’s restrictions on background Bluetooth scanning to preserve battery life.
• Usability (Demographics): The user base includes older members who may be unfamiliar with technology like QR scanning, requiring the design to be highly intuitive with alternative check-in methods

Engineering Standards

• Encryption Standards (TLS 1.3 / AES-256): Enforced by IETF / NIST. These are cryptographic protocols for secure communication and data storage. The project uses TLS 1.3 for all data in transit and AES-256 for data at rest in the database to protect PII.
• JSON Web Token (RFC 7519): Enforced by IETF. This is a compact, URL-safe means of representing claims to be transferred between two parties. The system uses JWT for secure, stateless user authentication with refresh token rotation.
• Bluetooth Low Energy (iBeacon Protocol): Enforced by Bluetooth SIG / Apple. This is a protocol for broadcasting identifier information to nearby electronic devices. The hardware and software adhere to this standard for the proximity detection system.
• GDPR Compliance (Privacy): Enforced by the European Union (adopted as a best practice). This is a regulation on data protection and privacy. The system complies by offering a “right to be forgotten” via data deletion requests and configurable data retention policies.
• RESTful API Architecture: Enforced by Industry Best Practices (Fielding). This is a software architectural style that defines a set of constraints to be used for creating Web services. The backend is designed with clean, versioned REST endpoints (e.g., /api/v1/members)

System Overview

The ROMAP system relies on a multi-layered architecture. The Presentation Layer consists of
a React Native mobile application supporting both iOS and Android, handling user interactions,
QR scanning, and background BLE detection. The Application Layer is a Node.js backend
using Express.js, which manages authentication, business logic, and API endpoints. The Data
Layer utilizes a PostgreSQL database managed via Prisma ORM for type-safe database
access. A unique Automation Layer employs a service combining Playwright and GPT-4 Vision
to capture screenshots of the external district portal, identify fields visually, and populate data
automatically. Finally, the IoT Layer integrates Kontakt.io BLE beacons configured to broadcast
specific UUIDs for meeting location validation.

Results

The pilot deployment of ROMAP resulted in a 95% reduction in average check-in time (from 3-5
minutes to 15 seconds) and the complete elimination of manual data entry errors. The BLE
proximity system achieved a 97% detection accuracy within a 10-meter range, and the AI
automation module successfully handled 94.4% of submission batches, showing resilience
against portal UI updates.

Screenshot Gallery

Future Work

Short-term goals include implementing multi-club federation support and an advanced analytics
dashboard. Medium-term enhancements involve automating grant reports using GPT-4 Vision
and integrating with wearable devices like Apple Watch. Long-term vision includes direct
integration with Rotary International systems if APIs become available and exploring blockchain
for immutable service records

Project Files

Project Charter
System Requirements Specification
Architectural Design Specification
Detailed Design Specification
Poster
Closeout Materials

References

[1] BlueBits Academy. (2021, June 1). React Native CLI: Getting started [Video]. YouTube.
[2] Chen, Y., Liu, Y., & Zhang, Y. (2019). BLE-based attendance system for large-scale
classrooms.
[3] Dan’s React Native Lab. (2022, November 7). React-Native and Bluetooth Low Energy (BLE)
[Video]. YouTube.
[4] IEEE. (2000). Design of an automated data entry system for hand-filled forms.
[5] Fielding, R. T. (2000). Architectural styles and the design of network-based software
architectures.
[7] Lamothe, M., et al. (2021). A Systematic Review of API Evolution Literature