AI/ML-Based Decision Support System for Intelligent Rake Formation and Logistics Optimization Across Sail Steel Plants

Abstract

The Steel Authority of India Limited (SAIL) operates multiple integrated steel plants across India, each generating and consuming large volumes of raw materials and finished products that necessitate efficient rail-based logistics. Rake formation—the process of assembling, routing, and dispatching railway wagons—remains a largely experience-driven, manual task prone to suboptimal utilization and scheduling conflicts. This paper presents the design and development of an AI/ML-Based Decision Support System (DSS) for intelligent rake formation and logistics optimization across SAIL steel plants. The proposed system integrates historical dispatch data, real-time wagon availability, plant production schedules, and Indian Railways network constraints into a unified analytics framework. Machine learning models including Random Forest, Gradient Boosting, and Long Short-Term Memory (LSTM) networks are employed to forecast demand, classify wagon requirements, and optimize rake assembly sequences. A multi-objective optimization engine balances throughput, turnaround time, and cost minimization under dynamic operating constraints. The DSS delivers actionable recommendations to logistics planners through an intuitive web-based dashboard. Evaluation on six months of operational data from SAIL’s Bhilai, Bokaro, Rourkela, Durgapur, and IISCO plants demonstrates a 23% improvement in wagon utilization, a 31% reduction in average rake turnaround time, and estimated annual logistics savings exceeding ₹180 crore across the network.