LIVE DEMO • SIH 2024 PROJECT

AI-Powered Railway
Traffic Intelligence

Real-time conflict detection, precedence optimization, and what-if simulation for Indian Railways section controllers. Built with OR-Tools CP-SAT + deep learning.

Live Demo →View on GitHub
Trains Managed Daily
Avg Delay Reduction
System Uptime
CAPABILITIES

Everything a Section Controller Needs

From real-time monitoring to advanced AI optimization, RailTrack AI covers every aspect of railway traffic management.

Real-Time Conflict Detection

AI monitors every train movement and predicts crossing conflicts seconds before they occur, giving controllers actionable time to act.

🧮

What-If Simulation

Simulate disruptions before they happen. Run scenarios with delays, breakdowns, or weather events and compare optimized outcomes side-by-side.

📊

Predictive Delay Engine

XGBoost + LSTM models trained on historical patterns forecast delays up to 2 hours ahead with station-level granularity.

👥

Multi-Role Access Control

Purpose-built dashboards for Section Controllers, Traffic Supervisors, Logistics Operators, and Admins with role-scoped data.

📋

Immutable Audit Trail

Every AI recommendation, override, and manual decision is logged with operator ID, timestamp, and outcome — full regulatory compliance.

🛰️

Real-Time IoT Integration

Kafka-powered event stream ingests signalling, rolling stock telemetry, and weather APIs in real-time with sub-second latency.

SYSTEM ARCHITECTURE

Production-Grade Infrastructure

Next.js Frontend
FastAPI Backend
OR-Tools Solver
ML Inference
PostgreSQL + TimescaleDB
Redis + Kafka
IoT Sensors → Kafka → ML Pipeline → OR-Tools → WebSocket → Frontend
BUILT WITH
Next.js 14FastAPIPostgreSQLTimescaleDBRedisApache KafkaOR-ToolsPyTorchDockerTailwind CSS