Port Downtime & Freight Delay Prediction

Problem

PSA Code Sprint 2024 set the theme “Resilient Port Operations Globally,” and global trade is nothing if not uncertain — unplanned downtime and freight delays cascade through supply chains and the cost lands on whoever sees them last. This project forecasts both from historical global port data so operators can plan ahead instead of reacting, contributing to PSA’s efficiency and sustainability goals.

Architecture

• A TensorFlow model predicts port downtime and freight delays from historical global port data.
• A Django backend handles data processing, serves the model, and manages data through the Django admin.
• A React dashboard visualizes predictions and insights.
• API integration accepts real-time inputs, so predictions update dynamically as new data arrives.

Built on Python 3.9+ with a React frontend and Django backend.

How it’s used

Pick a country’s port and freight to analyze; load data through the Django admin or the API; the model predicts downtime, and the dashboard renders it so teams can anticipate disruptions before they hit.