AI simulator optimises multi-warehouse inventory planning

Inventory planning across distributed warehouse networks remains a complex task as companies balance service levels, transport costs and fluctuating demand in modern digital supply chain operations. In this context, the MIT Center for Transportation & Logistics (MIT CTL) and Mecalux developed GENESIS, an AI-based simulator designed to optimise inventory distribution across logistics networks.

Simulating thousands of inventory strategies before execution
The Genetic Evaluation & Simulation for Inventory Strategy (GENESIS) platform uses machine learning models and genetic algorithms to evaluate thousands of inventory allocation scenarios. The objective is to determine optimal stock levels for each warehouse and identify when replenishment should take place.

The system evaluates factors including regional demand forecasts, transport costs and warehouse operational capacity. By simulating replenishment strategies in a virtual environment, companies can test logistics policies without affecting live operations.

Once operational data is entered, the platform generates optimisation recommendations supported by statistical dashboards. These include indicators such as consumption trends, demand variability by region, stock-keeping units (SKUs) with higher stockout risks and facilities experiencing supply constraints.

Rebalancing inventory instead of triggering new orders
A core function of the system is inventory rebalancing across warehouse networks. Rather than automatically triggering supplier orders, the platform evaluates whether transferring stock from another warehouse with surplus inventory would be more efficient.

This approach enables companies to reduce procurement and transport costs while improving utilisation of existing stock. The system also provides recommendations on transport planning, such as whether shipments should be consolidated to improve vehicle utilisation or dispatched from specific locations to reduce delivery times.




Faster scenario modelling for operational planning
The GENESIS platform was developed to evaluate multiple inventory strategies simultaneously rather than sequentially. This parallel simulation capability reduces analysis time from days to minutes, allowing the tool to support tactical logistics planning rather than only long-term analysis.

The platform is designed for use by both technical specialists and operational decision-makers, enabling broader access to simulation-based logistics optimisation tools.

Expanding AI use across warehouse operations
GENESIS represents one outcome of the collaboration between Mecalux and MIT CTL on AI applications in logistics. Further development work is focused on extending AI use cases to internal replenishment processes, digital twin models for high-density automated storage environments and slotting optimisation.

The collaboration reflects broader efforts to apply AI-based modelling tools to improve inventory visibility, warehouse automation strategies and logistics network performance.

Edited by industrial journalist, Aishwarya Mambet — AI-powered.

www.mecalux.com