Robots Atlas>ROBOTS ATLAS
Robotics

Warehouse automation

2012ActivePublished: 9 November 2026Updated: 9 November 2026Published
Key innovation
Replacing the "person-to-goods" model with a "goods-to-person" model and a coordinated robot fleet, decoupling warehouse throughput from the number of workers walking between racks.
Category
Robotics
Abstraction level
Paradigm
Operation level
Robot controlApplicationDeployment
Use cases
E-commerce order fulfillmentDistribution centers and high-bay warehousesCold storage and temperature-controlled warehousesCross-docking and transshipmentOrder picking and palletizingInventory management and stocktaking

How it works

A WMS/WES receives orders and plans the workflow. AMR/AGV robots and AS/RS systems (storage/retrieval machines, SRM) move goods or entire shelving units to picking stations (goods-to-person) or directly to dispatch points. Machine vision and sensors (LiDAR, cameras) enable navigation, localization (SLAM) and object recognition, while barcodes/RFID identify and track loads. Picking and palletizing robots handle individual items, sorters and conveyors route parcels to the right locations, and orchestration software coordinates the fleet, optimizes routes and updates inventory in real time.

Problem solved

Traditional picking based on workers walking between racks is slow, costly, error-prone and hard to scale with rising e-commerce volumes. Warehouse automation addresses throughput, accuracy, space utilization and labor availability.

Components

Autonomous mobile robots and automated guided vehicles (AMR/AGV)Internal transport and goods-to-person

Mobile platforms moving goods or whole shelving units to picking stations; AMRs navigate freely using SLAM, AGVs follow predefined paths.

Official

Automated storage and retrieval systems (AS/RS)High-density storage

Computer-controlled storage/retrieval machines (SRM), cranes, shuttle systems and vertical carousels providing dense, accurate storage and automated load retrieval.

Official

Picking and palletizing robotsSingle-item manipulation

Industrial arms (4–6 axis) and grippers for picking, packing, palletizing and depalletizing, often aided by machine vision for object recognition and grasping.

Official

Sorters and conveyorsFlow and parcel diverting

Conveyors and high-throughput sortation systems routing cartons and parcels to the correct locations or dispatch gates.

Official

WMS/WES/WCS softwareOrchestration and control

Warehouse management (WMS), execution (WES) and equipment control (WCS) software layer: order intake, stock assignment, fleet planning and coordination, and real-time inventory updates.

Official

Machine vision and identification (barcode/RFID)Perception and load tracking

Cameras and sensors for navigation, object recognition and quality control, plus load identification via barcodes and RFID tags.

Official

Evolution

Original paper · 2008 · AI Magazine (2008) / IAAI 2007 · Peter R. Wurman
Coordinating Hundreds of Cooperative, Autonomous Vehicles in Warehouses
Peter R. Wurman, Raffaello D'Andrea, Mick Mountz
1962
First AS/RS systems

In the 1960s the first automated storage and retrieval systems appeared, initially for heavy pallet loads.

2003
Kiva Systems and the goods-to-person model

Kiva Systems (originally Distrobot) is founded, developing robots that bring shelves to stationary workers rather than the reverse.

2012
Amazon acquires Kiva Systems
Inflection point

Amazon acquires Kiva Systems for about $775M (later Amazon Robotics), mainstreaming large-scale mobile warehouse robotics.

2021
Humanoids enter logistics

Agility Robotics' Digit humanoid begins being used for bulk material handling and logistics tasks in warehouses.

2025
Amazon deploys its one-millionth robot and the DeepFleet model

Amazon announces its one-millionth warehouse robot and DeepFleet, an AI model coordinating fleet routes, making the fleet about 10% faster.

Execution paradigm

Primary mode
Conditional

Event-driven operation: orders and inventory state dynamically trigger robot tasks and fleet routing.

Activation pattern
Input dependent