AI’s new role in running the warehouse

As global supply chains brace for another intense peak season, a new study from Mecalux and MIT’s Intelligent Logistics Systems Lab offers a look at the modern warehouse. It is a warehouse no longer defined by just racking, forklifts, and labor. It is increasingly defined by artificial intelligence.

The research, based on responses from more than 2,000 warehouse and supply chain leaders across 21 countries, shows that artificial intelligence and machine learning have moved from pilots to production systems, reshaping productivity, workforce strategy, and decision-making at an unprecedented scale.

The report noted that more than 90% of warehouses now use some form of AI or advanced automation, and roughly 60% are operating at advanced maturity levels.

AI maturity has arrived

Across industries, warehouses report that AI now supports day-to-day workflows including:

• Order picking and routing
• Inventory accuracy and slotting optimization
• Predictive maintenance
• Labor planning and performance monitoring
• Safety and ergonomic risk detection

This level of operational penetration marks a shift from experimentation to mainstream practicality.

“The data show that intelligent warehouses outperform not only in volume and accuracy, but in adaptability,” Mecalux CEO Javier Carrillo noted in the report. “As peak season approaches, companies that have invested in AI aren’t just faster, they’re more resilient, more predictable, and better positioned to navigate volatility.”

Those performance gains are showing up in hard numbers. The typical AI payback period is now two to three years, significantly faster than earlier automation investments, the report found. Companies attribute this ROI to:

• Higher inventory accuracy
• Reduction in picking errors
• Throughput increases
• Labor productivity gains
• Less unplanned equipment downtime

Many respondents now dedicate 11% to 30% of their warehouse technology budgets to AI initiatives.

The workforce is expanding

While many assume AI is leading to job losses, the report found something different happening. AI adoption is correlated with more hiring and higher worker satisfaction, it noted.

More than three-quarters of surveyed organizations saw a rise in employee satisfaction after implementing AI, and over half reported increased workforce size, driven by new roles such as AI/ML engineers, automation specialists, process-improvement experts, and data scientists.

The MIT team emphasized that automation is not eliminating frontline roles but elevating them. Workers spend less time on repetitive tasks and more time on oversight, troubleshooting, analytics, and exception management.

The “last mile” of AI integration

Even with widespread adoption, the report highlights meaningful barriers that keep organizations from fully realizing AI’s value. According to Dr. Matthias Winkenbach, director of the MIT Intelligent Logistics Systems Lab, challenges such as a lack of technical expertise, poor data quality, integration challenges with legacy WMS and ERP systems, high upfront costs, and issues scaling pilots to multiple sites are still hampering organizations.

“The hard part now is the last mile [of] integrating people, data, and analytics seamlessly into existing systems,” he said.

Still, many companies say they have strong foundational capabilities in project management and data governance. What they need now are clearer roadmaps and increased budgets, both of which respondents expect to expand over the next two to three years.

Generative AI emerges as a breakthrough driver

The study identifies generative AI as the single most valuable AI method used in today’s warehouses, outpacing predictive analytics and computer vision in perceived impact. Among its most promising applications are automated documentation and labeling, code generation for automation systems, warehouse layout design, process-flow optimization, and knowledge capture and task guidance.

“Traditional machine learning is great at predicting problems, but generative AI actually helps you engineer the solution,” Winkenbach noted.

This shift suggests a new era where AI not only optimizes decisions but designs operational improvements autonomously.

According to the study, 87% of companies plan to increase AI budgets and 92% have new AI projects underway. That momentum signals a future in which warehouses become increasingly self-optimizing. Over the next few years, the report sees greater use of multimodal AI to merge video, sensor, and operational data, and autonomous maintenance and self-correcting systems. Further, an increased integration between AI-driven labor planning and robotics and a convergence of generative AI with real-time execution systems should drive future productivity.

The result, the report argues, will be warehouse networks that are not only automated but intelligent, adaptive, and highly resilient.

As logistics organizations push into 2026, the MIT–Mecalux study makes one conclusion clear: AI is no longer a pilot program. It is a competitive requirement.

Warehouses that integrate AI into daily operations that are supported by better data, stronger workforce development, and cross-functional implementation, will outperform not just in speed, but in the resilience and adaptability that define modern supply chains.

SC
MR