ODSP2115 Pre-recorded On demand
AI-supported decision making in transport: Optimizing split orders
Hassan Rashid, Kamil Zukowski
Summary
From Microsoft's AI summary
EPAM and Albert Heijn present a proof of concept using AI and Microsoft Foundry to help transport planners split overloaded truck orders. The system is advisory, keeping a human in the loop, and combines a shared business ontology, constraints, ranking, route validation and AI reasoning. It has a strict rules mode and a relaxed mode. It was built with Azure, .NET, Microsoft Agent Framework and Azure OpenAI. A demo shows recommendations with explanations. The speakers reported higher truck fill rates, fewer vehicles and faster operations. No availability or pricing was stated.
Session description
Microsoft's catalogue description
When delivery routes exceed truck capacity, planners make rapid decisions that affect cost, service levels, and operational efficiency. EPAM and Ahold Delhaize share how they captured expert planner knowledge and turned it into an AI-powered decision-support solution built on Microsoft AI. Learn how AI, real-time optimization, and human-in-the-loop decision making reduced standby truck usage, improved route utilization, and created a foundation for intelligent supply chain operations.