Problem:

SignalGuard was a collaborative project developed by students from different IT specializations, including Application Development, AI & Data Visualization, Cloud & Cybersecurity, and Digital Innovation. The project was built for Cipal Schaubroeck, a company that specializes in ANPR camera systems for traffic monitoring. Since these cameras were a relatively new addition to their product line, Cipal needed a monitoring system to help both their technicians and clients identify and resolve issues faster.

Understanding the Problem:

Our project focused on solving three major challenges:

Remote Power Cycling: Many ANPR camera issues could be resolved simply by powering the system off and on again, yet technicians still had to be sent on-site to do this manually. We aimed to create a solution that would allow remote resets with the press of a button.
Automated Alert Prioritization: The system generated an overwhelming number of error messages and status updates every hour. These alerts had to be filtered manually, which was inefficient. We sought to automate the prioritization of these messages, ensuring that only the most critical issues required immediate attention.
Last Line of Communication: When a complete system failure occurred, there was no fallback method of communication. This meant that technicians were sent in blindly, increasing the risk of delayed repairs and putting Cipal at risk of breaching Service Level Agreements (SLAs) with their clients.

Project Roles & Collaboration:

Each specialization within our team focused on a different part of the system:

Application Development students and I worked on building the frontend and backend of the application.
The AI & Data Visualization student analyzed system data and worked on automated alert prioritization.
The Cloud & Cybersecurity student ensured that the system was securely hosted and deployed online.
As a Digital Innovation student, I took on the role of project manager, leading the daily scrum meetings and ensuring effective collaboration between all team members.

Technical Approach & Implementation:

For the frontend, we used React, while the backend was built using Fastify, a modern web framework for Node.js that outperforms ExpressJS in terms of speed and efficiency. We chose PostgreSQL as our database and integrated Prisma ORM for secure and optimized SQL operations.

On the hardware side, we deployed an OrangePI mini-computer inside the ANPR camera system's electrical box. This allowed us to monitor power levels and breaker switches, ensuring that power loss issues could be detected remotely. We also installed a relay board, which allowed us to remotely cut and restore power to critical components, such as the internet modem.

Due to the tight three-week development timeline, we focused on building a proof of concept (PoC) with core functionality. Unfortunately, we did not have enough time to implement real-time data processing or full logging capabilities.

Challenges & Limitations:

Learning Curve with Fastify: Fastify provided performance benefits over older frameworks, but its lack of documentation made it difficult to work with at times.
Lack of Quality Data for AI: We initially planned to integrate AI-driven issue detection, but the available data was too unstructured to train a reliable model. As a result, AI features were ultimately scrapped from the project.
Time Constraints: With only three sprints (one week each) to develop the project, we had to prioritize essential features. This meant that advanced functionalities, such as real-time monitoring and data visualization, had to be left out.

My Contributions & Learning Experience:

In addition to leading project management efforts, I personally developed the JWT-based authentication system for the frontend and backend, including full role-based access control. I also helped with feature planning and UI design, ensuring that the application included:

A map-based interface displaying all ANPR camera locations
A notification dashboard for tracking alerts
A client management system

Even though the AI component was not implemented, I was actively involved in researching possible AI integrations and learned a lot about data structuring and analysis during the process.

Project Outcome & Future Potential:

The final product was delivered to Cipal Schaubroeck as a Proof of Concept (PoC). While the company is unlikely to use our exact implementation, our work laid the foundation for their future monitoring system. Certain features—like remote power cycling—could potentially be integrated into their production systems, helping reduce maintenance costs and improve response times.

Even though this project had its limitations, it was a valuable learning experience, showing me the importance of collaborating with different IT disciplines, prioritizing tasks under time pressure, and adapting to unexpected challenges.

Links:

GitLab

César Van Leuffelen

7 projects

MobiLAB & Care REV app

LABflow

Thomas More Guide

Grow More

SignalGuard - Project 4.0

PURplus Planning App

Blended Intensive Program