OceanGuard Analytics

OceanGuard Analytics — Maritime Accident Prevention Intelligence Platform

📖 Overview

OceanGuard Analytics is a live, open-access maritime accident prevention platform built by a Marine Engineer turned Data Professional, independently, from Lagos, Nigeria, with zero institutional funding.

While the MAIB Maritime Accident Analysis project uncovered the hidden patterns in five years of UK government accident data, OceanGuard is what those patterns become when you put them in the hands of the people who need them most — port authority officers, coastguard commanders, maritime regulators and environmental NGOs across Africa.

This is not another analysis project. This is a deployed prevention tool.

The core question OceanGuard answers: Which vessel types, location zones, weather conditions, and seasonal patterns are most likely to produce the next maritime accident  and the next pollution event — before it happens?

🛠 Tools Used

• Python 

• Pandas  

• Plotly 

• Streamlit 

• Power BI 

•  GitHub 

• Streamlit Cloud

📊 Platform Features

✅ 7 KPI Indicators — Total accidents, fatalities, fatal incidents, ships lost, top vessel type, top injury, top location zone — visible in under 5 seconds

✅ Accidents by Year — Trend analysis across 2020 to 2024 showing year-over-year patterns

✅ Vessel Type Breakdown — Donut chart revealing which vessel categories dominate accident records

✅ Injury Intelligence — Top 8 injury types by frequency, directly informing safety training priorities

✅ Weather Correlation — Counterintuitive finding: clear weather produces the most accidents, not storms

✅ Location Zone Analysis — Horizontal bar chart mapping accident frequency across coastal, port, open sea, and inland zones

✅ Accident Heatmap — Vessel type vs year risk matrix — the most powerful single view for fleet inspectors and regulators

✅ Damage Assessment — Distribution from no damage through to total ship loss

✅ Searchable Accident Log — Free-text search across all 1,840 verified records

✅ Multi-Country Architecture — Country filter built in and ready to integrate Nigerian NIMASA data, EMSA European records, and USCG data

🔑 Key Insights

1. Passenger ships dominate incidents but not fatalities With 758 incidents, passenger ships are the most accident-prone vessel type but their fatality rate is remarkably low, evidence of strong ISM and STCW compliance. The prevention implication: focus on port congestion and navigation aids, not crew training.

2. Recreational craft are the deadliest vessel type.  Despite fewer total incidents, recreational craft account for the highest fatalities, operators without professional oversight, limited safety equipment, and no mandatory weather monitoring. A targeted public education campaign could prevent the majority of these deaths.

3. Clear weather is the most dangerous condition 46 to 50% of all accidents occur in clear weather, not storms, not fog. Overconfidence in good visibility drives human-error collisions. This single finding challenges conventional maritime safety assumptions.

4. Port areas are the highest-risk location zone Internal waters and port areas account for the highest accident frequency,  dense traffic, complex manoeuvring, and human fatigue converge in the same space.

5. Prevention beats reaction 82 fatal incidents over 4 years are entirely preventable. The patterns are clear. The data exists. What was missing was a platform that makes those patterns actionable in real time, that is what OceanGuard provides.

🌍 Why Africa

Africa has 38 coastal nations, over 40,000 km of coastline, and some of the world's most biodiverse marine ecosystems. Almost none have structured maritime accident intelligence systems. Accident data sits in PDF reports, unstructured and inaccessible to the officers who need it.

OceanGuard is built to change that, starting with Nigeria, replicable across the continent. The multi-country architecture is already live. The methodology is open-source and free. Any African maritime authority can adopt it using their own national data.

🔮 Future Roadmap

Phase 1 — Nigeria Data Integration Digitize and structure NIMASA accident records creating the first comparative West African maritime safety dataset

Phase 2 — Multi-Country Expansion Add EMSA (Europe), USCG (USA) and Transport Canada data, enabling the first cross-continental maritime safety comparison

Phase 3 — Predictive Risk Scoring ML model flagging high-risk vessel-route-season combinations before accidents occur,  building on the existing Predictive Maintenance for Marine Engines project

Phase 4 — Open Replication Toolkit Free methodology guide enabling Ghana, Kenya, Tanzania and other African coastal nations to build their own national instances at near-zero cost

📬 Live Links

🔗 Live OceanGuard Platform — explore the full dashboard

💻 GitHub Repository — open-source code, fully documented

📊 MAIB Source Analysis — the research project that powers this platform

📬 Let's Connect

Have ideas, feedback, or want to collaborate?

📧 fijaytwo@gmail.com

🔗 LinkedIn