Researchers at the MIT Center for Transportation & Logistics and intralogistics company Mecalux have developed a new artificial intelligence–powered simulator designed to help companies optimize how inventory is distributed across multiple warehouses within a logistics network.

Read also: 8 Automation Technologies Reshaping the Modern Warehouse in 2026

The platform, known as Genetic Evaluation & Simulation for Inventory Strategy (GENESIS), was created by the Intelligent Logistics Systems Lab (ILS) and uses advanced machine learning and genetic algorithms to evaluate thousands of operational scenarios. The system determines the most efficient inventory levels for each warehouse while also recommending optimal replenishment strategies.

GENESIS analyzes a wide range of operational variables, including regional demand forecasts, transportation costs, and warehouse capacity. By simulating different inventory policies in a virtual environment, companies can test strategies without disrupting their real-world operations.

Matthias Winkenbach, Director of Research at the MIT Center for Transportation & Logistics and the Intelligent Logistics Systems Lab, said the system’s genetic algorithm enables rapid evaluation of multiple logistics scenarios.

According to Winkenbach, the simulator allows companies to compare various strategies and identify the approach that best aligns with their operational needs.

Once operational data and parameters are entered into the system, GENESIS produces an optimized inventory plan supported by analytical dashboards. These dashboards highlight key metrics such as consumption trends, regions with high demand volatility, SKUs at risk of stockouts, and warehouses experiencing supply imbalances.

Redistributing Inventory Before Reordering

A key feature of the platform is its ability to rebalance inventory within a logistics network. Instead of immediately triggering new purchase orders, the system evaluates whether excess stock from one facility can be transferred to another warehouse facing shortages.

This approach helps companies reduce procurement costs while improving the use of existing inventory.

The simulator also provides guidance on transportation strategies. It can recommend consolidating shipments to maximize truck capacity or fulfilling orders from specific warehouses to reduce delivery times and transportation expenses.

Rodrigo Hermosilla, a research engineer at the MIT Intelligent Logistics Systems Lab, explained that the main challenge in developing the system was not identifying the right algorithm but making the technology fast enough for practical use.

By designing GENESIS to evaluate thousands of scenarios simultaneously rather than sequentially, the team significantly reduced processing time. What previously required days of analysis can now be completed in minutes, making the platform suitable for real-world tactical planning.

Unlike many advanced analytics tools that require specialized technical expertise, GENESIS was designed to be accessible to both data specialists and business decision-makers.

Javier Carrillo, CEO of Mecalux, said the platform aims to help companies minimize total logistics network costs while maintaining high service levels.

Expanding AI in Logistics

The AI-based simulator represents one of the first major outcomes of the collaboration between Mecalux and the Massachusetts Institute of Technology logistics research community.

The partnership is now moving into a new phase focused on expanding artificial intelligence applications across other logistics processes. Planned initiatives include AI-driven internal replenishment systems, digital twin technology for high-density automated storage facilities, and advanced slotting optimization tools.

As supply chains grow increasingly complex, researchers and industry partners say tools like GENESIS could play a key role in helping companies make faster, data-driven decisions while reducing operational costs.