Russian University Develops AI Warehouse Management System for Major Retailers
A new AI system designed by researchers in Moscow aims to optimize inventory decisions and improve profitability for large retail distributors.
Researchers at the Institut iskusstvennogo intellekta MFTI (AI Institute of the Moscow Institute of Physics and Technology) have developed an AI-based warehouse management system for large retail companies, the university’s press service said.
The development addresses one of the most resource-intensive tasks in logistics—inventory management. While standard auto-replenishment systems typically generate a demand forecast and then apply a fixed formula, the RL-agent created by the Moscow-based researchers optimizes decisions directly around a target business metric, taking into account the long-term consequences of each order.
In tests using historical data, the system increased the share of fulfilled demand from 80% to 90%, directly reducing lost revenue. Another key feature is that the agent adapts to changes in logistics parameters and market conditions without requiring manual reconfiguration.
A Digital Twin for Testing
The team also created a digital copy of business processes—a platform that replays sales history and allows researchers to test different strategies in conditions as close to real-world operations as possible.
