Amazon deploys 1M warehouse robots with new AI traffic controller

Amazon has reached a major milestone with over one million warehouse robots deployed across more than 300 fulfillment centers, while unveiling DeepFleet, a new AI foundation model that acts as a traffic controller for its robotic fleet. The achievement marks Amazon’s evolution from testing a handful of shelf-moving robots in 2012 to operating a massive automated workforce that the company says creates more jobs rather than eliminating them.

The big picture: Amazon’s robotics strategy centers on helping human workers access inventory more efficiently rather than replacing them entirely.

• CEO Andy Jassy recently tweeted the company’s north star question: “How can we help operations employees access inventory more efficiently?”
• The robots are squat carts the size of a footstool that slide under shelving towers, lift entire racks, and bring them to workers who pick ordered items.
• Every trip shaves walking time from human packers’ shifts and pushes orders closer to shipping.

How DeepFleet works: The new AI system uses live warehouse data to optimize robot movements in real-time, functioning like an intelligent traffic management system.

• Built on Amazon’s AWS infrastructure, DeepFleet reroutes robots on the fly to avoid jams and bottlenecks.
• “Think of DeepFleet as an intelligent traffic management system for a city filled with cars moving through congested streets,” says Scott Dresser, VP of Amazon Robotics.
• The AI model learns continuously to steer robots around slow spots, cutting travel time by about 10% and speeding up customer orders.

Job creation claims: Amazon argues that advanced robotics create more positions rather than eliminating them, particularly in technical roles.

• “Advanced robotics require 30% more employees in reliability, maintenance, and engineering roles,” Dresser said.
• New positions range from motor repair technicians to data analysts who track robot movement patterns to identify delays.
• At Amazon’s Shreveport, Louisiana fulfillment center, which opened last year with the latest robot generation, the company says robots created new jobs rather than replacing existing ones.

Worker training initiatives: Since 2019, more than 700,000 Amazon workers have participated in programs to develop technical skills for an increasingly automated economy.

• The company says this experience prepares workers for jobs whether they stay at Amazon or move to other employers.
• Training focuses on maintaining and improving automated systems rather than just operating them.

Competitive landscape: Amazon isn’t alone in warehouse automation, with several major retailers investing heavily in robotic systems.

• U.K.-based Ocado uses thousands of compact robots in grocery warehouses that zip across grid-like platforms, picking items with near-perfect accuracy.
• Walmart is working with Symbotic to bring robot-powered sorting to 42 distribution centers by 2030, with CasePick robots that can sort cases in under a minute.

What’s next: Amazon plans to leverage DeepFleet data to reduce energy consumption and experiment with smaller fulfillment centers closer to urban areas.

• The company is testing “micro-fulfillment” centers that could enable faster delivery with a smaller carbon footprint.
• Future AI systems won’t just direct traffic but will help decide what products to stock and when, creating warehouses that adjust themselves in real-time.

What they’re saying: Amazon executives see this as just the beginning of AI-driven logistics transformation.

• “This is just the beginning. As DeepFleet learns from more data, it will continue to get smarter, driving deeper efficiencies, unlocking more selection closer to customers, and reimagining what’s possible in robotic logistics,” Dresser says.
• The company acknowledges uncertainty about how people will fit into a world where machines not only move shelves but make decisions too.