"Accelerating learning and algorithms for logistics problems."
The Algorithms and Machine Learning (ML) vertical at the IIT Madras-led FedEx SMART Center is committed to advancing logistics and supply chain operations through advanced computational techniques. The focus is on building scalable, intelligent solutions to manage complex challenges in logistics.
Some of the projects under this vertical include:
Reinforcement Learning for train dispatching and re-routing.
Reinforcement learning to arrange convex and non-convex objects.
Quantum Machine Learning implementations.
Scalable solutions to logistics problems (CVRP- Capacitated Vehicle Routing Problem) using parallelization.
Key focus areas include:
Predictive Modeling and Forecasting - Forecasting demand, staffing, capacity, and shipment allocation to optimize operational planning.
Optimization Algorithms- Designing scalable, high-performance algorithms to solve complex problems central to logistics, such as the Capacitated Vehicle Routing Problem (CVRP), 3D bin packing, and scheduling. Techniques include heuristic and metaheuristic approaches, parallel computing, and hybrid solvers to ensure fast, near-optimal solutions at scale. The goal is to enable smarter resource utilization, reduced costs, and improved operational efficiency across logistics networks.
Reinforcement Learning for Logistics : Explores the application of reinforcement learning (RL) to dynamic, real-time decision-making in logistics operations.
Quantum Computing Applications - Exploring hybrid quantum-classical approaches to accelerate solutions for intensive tasks like routing and packing.
Expected outcomes include the optimization of routes, resource allocation, and container utilization to enhance overall operational efficiency, while improving customer experience through accurate demand forecasting and real-time shipment tracking.
