Maritim Guard

Real-Time Protection, Real-World Impact

Comment

Inspiration

Data from KNKT shows that more than 190 major maritime accidents occurred in Indonesia between 2015 and 2025, claiming more than 787 lives. Recurring patterns in these cases include old vessels, overloading, inaccurate manifests, insufficient equipment and implementation of standard operating procedures (SOPs) for safety, as well as weak supervision at departure points.

What it does

MaritimeGuard provides real-time corrosion monitoring for ships through a smart, sensor-integrated web platform. It continuously tracks hull conditions using embedded sensors that measure pH, temperature, humidity, and conductivity. This data is transmitted live to a centralized dashboard where operators can visualize the status of each vessel in the fleet. The platform uses color-coded indicators to highlight ships at risk and offers detailed sensor readings per ship. It also includes AI-generated summaries that highlight trends, potential threats, and recommended actions making it easier for fleet managers to make informed decisions. When corrosion thresholds are exceeded, MaritimeGuard automatically sends alerts, helping crews prioritize inspections and maintenance before serious damage occurs. The system also supports compliance efforts by generating ready to download reports that align with maritime regulatory standards, making inspections and audits more efficient.

How we built it

We divided the project into two main parts to streamline development: Smart Devices and Web Platform.

  1. Smart Devices We designed and built custom IoT hardware focused on predictive maintenance for ship hulls. The system monitors four key parameters to assess corrosion risks:
  2. Humidity (sensor)
  3. pH level (sensor)
  4. Electrical conductivity (sensor)
  5. Dew point (calculated from humidity and temperature)
  6. These values are processed locally and then transmitted to the cloud for live monitoring and alerting.

  7. Our goal was to create a smart, field-ready device that could operate reliably in harsh maritime environments.

  8. Web Platform To support and visualize sensor data, we built a full-stack system using modern web technologies:

Redis Pub/Sub was used for efficient, real-time data communication between IoT devices and the web dashboard.

Next.js handled both the frontend (UI, dashboard, landing page) and backend (data processing, API routes), allowing for scalable and cohesive development.

Deployment was split:

Frontend was hosted on Vercel for fast global delivery

Backend and message processing ran on a VPS for better control over server resources

This architecture allows us to monitor multiple ships in real-time while providing a responsive and intuitive interface for fleet operators.

Challenges we ran into

One of the biggest challenges was ensuring real-time data accuracy and reliability from the sensors, especially considering the harsh and varying conditions at sea. Sensor calibration across different environments proved tricky, as each ship can have unique hull materials and exposure levels. Handling data latency in the dashboard was another obstacle. Since the system relies on live readings, even minor delays or drops in data transmission could affect the user's ability to act quickly. We had to optimize both the backend pipeline and frontend rendering logic to keep things smooth. Building an AI powered summary that turns raw sensor data into meaningful fleet insights was more complex than expected. Translating numeric trends into understandable language, while keeping it technically accurate, required fine-tuning and testing. We also faced UX design challenges finding the balance between providing detailed engineering data for maintenance crews and clear visual indicators for ship operators who might not have a technical background. Finally, making sure the platform’s reporting tools aligned with maritime compliance standards (ASTM or SNI) involved research and iterations to ensure everything was formatted and structured correctly.

Accomplishments that we're proud of

We’re proud to have built a working prototype of MaritimeGuard that successfully monitors and visualizes corrosion data in real time. Seeing live sensor readings appear on the dashboard and having them trigger alerts based on actual risk levels was a major milestone. We managed to integrate multiple components into a single cohesive system: hardware sensors, data streaming, backend processing, frontend visualization, and even AI-powered summaries. Bridging these different layers was both technically and logistically challenging, and it paid off. Our team also designed a clean and intuitive user interface that makes complex sensor data easy to understand for both technical and non technical users. The traffic light style indicators and detailed ship views are simple yet powerful. Another point of pride is our commitment to regulatory awareness designing the system with ASTM and SNI compliance in mind, which adds real-world applicability and value. Finally, we’re proud that our project directly addresses a real, high stakes problem in the maritime industry, with potential to reduce repair costs, improve ship safety, and modernize fleet maintenance.

What we learned

Throughout the development of MaritimeGuard, we learned how critical it is to bridge the gap between hardware and software especially in real-time systems. Working with live sensor data taught us about the importance of data validation, synchronization, and fault tolerance in unstable environments like the open sea. We gained a deeper understanding of corrosion science and how environmental factors like pH, conductivity, and humidity interact. This helped us design more meaningful risk indicators and alert logic. Designing a user-friendly dashboard taught us that clarity is just as important as accuracy. Complex sensor data must be presented in a way that allows quick, confident decision making even by users who aren’t engineers. Integrating AI summarization made us realize the challenge of turning raw data into insightful, human-readable output especially when the stakes involve asset integrity and safety. Lastly, we learned that solving a real-world problem, especially one that involves safety and compliance, requires cross-disciplinary thinking from maritime standards to UX, from machine learning to mechanical systems.

What's next for Maritim Guard

We see MaritimeGuard not just as a project, but as a growing solution with real-world impact. Here's where we're headed next:

  1. We're currently using simulated sensor data based on research journals about hull corrosion. The next step is deploying our system in real maritime environments to validate performance and improve reliability in field conditions.
  2. We plan to integrate Large Language Models (LLMs) into our platform to act as an intelligent assistant. This AI will help users interpret data, suggest actions, and even answer questions about corrosion risks, maintenance steps, and compliance requirements.
  3. To ensure long-term sustainability, we're developing a subscription-based model. This will support continuous development and allow us to offer scalable solutions for different fleet sizes and needs.
  4. Our bigger mission is to solve the problem of ship corrosion starting in Indonesia, a country with thousands of vessels. From there, we aim to expand globally bringing our technology to fleets around the world and making maritime operations safer, smarter, and more efficient.

Built With

Share this project:

Updates