Optimization AI is the decision-making engine of the warehouse. While optimization has existed for years through math-based models, AI introduces adaptability and predictive power.

“Even pre-AI, people have been using math to make decisions in warehouses,” Mabe said. “Where AI comes in is adding better inputs and adaptivity, really predicting velocity and learning how to restock inventory for changing demands.”

With supervised learning, systems can forecast SKU velocity and demand patterns. Reinforcement learning allows systems to respond to dynamic conditions like flash sales or seasonal shifts. The result is real-time optimization that delivers measurable ROI.

Vision and Perception AI use cameras and sensors to monitor warehouse activity, detect errors, and ensure safety. These technologies are essential for the future of autonomous warehouse operations.

“Vision is foundational for getting to the concept of a lights-out warehouse,” Mabe noted. “It’s about using cameras and sensors to process information and make decisions in real time.”

From spotting misaligned totes to identifying ergonomic risks, these systems help reduce downtime and improve operational flow, especially in highly automated environments.

Generative AI and Large Language Models (LLMs) act as the connective tissue of the warehouse, turning complex data into simple, actionable insights. Instead of navigating dashboards or spreadsheets, teams can ask straightforward questions and get immediate answers.

“You’ll be able to ask questions and get answers in seconds — charts, trends, whatever format is appropriate,” Mabe said. “It dramatically reduces the time from question to insights.”

This capability is already being piloted and is expected to scale quickly, helping teams solve problems faster and make better decisions on the fly.