Back

While trying to land an internship in a company, I decided to invest a few time building a fun idea I had on my mind which is a real-time maritime intelligence system that tracks commercial vessels arriving at major national ports in my home country and forecasts growth to potentially generate a signal before earnings.

The system pulls live vessel positions through an AIS data feed, cross-referencing each arrival against two satellites from the European Space Agency’s Copernicus programme one optical satellite for high-resolution yard imaging, and one SAR radar satellite that operates day and night through any weather.

The radar satellite then transmits microwave pulses across two polarisation channels and measures the energy reflected back each commodity produces a distinct backscatter signature based on its geometric structure and material conductivity, allowing the system to estimate container density and infer cargo type without ever seeing a manifest. 

Before the satellite even looks at the yard, a confidence scoring engine uses deadweight tonnage, gross tonnage, and draught readings to estimate shipment weight then tracks draught changes post-arrival to confirm what was actually offloaded.

When yard occupancy remains elevated beyond the free storage window, the shipment is flagged as likely incurring storage fees, which feeds directly into a port growth metric as a signal of sustained import activity.

Beyond individual shipments, the system continuously aggregates vessel arrivals, dwell times, and satellite-confirmed yard occupancy across all monitored ports building a forecast of port throughput and trade growth that leads official statistics by weeks.

The entire system runs on free and open-source data. Paid AIS feeds, commercial SAR tasking, and higher-frequency satellite revisits would significantly increase accuracy and detection confidence.

I promised myself I’d never do the LinkedIn project post thing but here we are with 5 connections let’s see how things goes!