Kiva Robots: How the Amazon Warehouse System Works

What Kiva Robots Were and Who Built Them

Kiva robots were orange, low-profile automated guided vehicles (AGVs) designed to move inventory across a distribution center floor. Mick Mountz, Peter Wurman, and Raffaello D'Andrea founded Kiva Systems in the early 2000s to solve a central problem in order fulfillment: the time human pickers spend walking warehouse aisles to retrieve individual items.

Instead of sending workers into aisles of static shelving, Kiva inverted the process into a goods-to-person model. The drive units traveled underneath portable storage pods, lifted them, and moved them across the building to a human worker at a station.

At The Bot Scout, we look at automation platforms through the lens of what operators can realistically deploy, and Kiva is the reference point most warehouse buyers start from.

How the Goods-to-Person Mechanism Moved Inventory

The core mechanism of a Kiva installation relied on three parts: mobile drive units, portable storage pods, and centralized control software. When an incoming customer order reached the facility management system, the software located the nearest available drive unit and dispatched it to the pod holding the target stock keeping unit (SKU).

The robots navigated by reading barcode stickers affixed to the warehouse floor on the warehouse floor and used sensors to detect obstacles and avoid collisions with other units or personnel.

Once the drive unit arrived underneath the appropriate pod, it lifted the shelf structure off the ground. The machine then carried the entire pod along designated transit paths to a human worker waiting at a stationary workstation, turning a long walking task into a stationary pick-and-pack workflow.

Original Unit Specifications Reported at Acquisition

Historical reporting from the era of the acquisition provides concrete operating parameters for early commercial models. Kiva manufactured two primary hardware variants to handle different load profiles within distribution facilities.

The smaller standard model measured roughly 2 by 2.5 feet, stood 18 inches high, and carried loads up to 1,000 pounds. A larger heavy-duty variant was engineered to transport pallet loads weighing up to 3,000 pounds across the facility.

Both drive models operated at a top travel speed of approximately 1.3 meters per second. Battery management relied on short, frequent duty cycles, with machines taking a five-minute recharge roughly every hour to keep the fleet circulating continuously without extended downtime.

The 2012 Amazon Acquisition and Fleet Expansion

In March 2012, Amazon announced the acquisition of Kiva Systems for $775 million in cash, which stood as Amazon's second-largest corporate acquisition at that time. Following the transaction, Amazon stopped marketing Kiva systems to outside customers and stopped announcing new commercial clients.

The corporate entity was officially renamed Amazon Robotics in August 2015. If you want to understand the modern internal setup, our guide to Amazon Robotics details how those operations look inside company fulfillment centers today.

Internal deployment expanded steadily after the sale. By June 2019, Amazon reported operating more than 200,000 mobile robots across its fulfillment network.

From Guided Drive Units to Autonomous Mobile Robots

In July 2022, Amazon unveiled Proteus, described as its first autonomous mobile robot (AMR).

Kiva-style hardware navigated by floor barcodes, and autonomous mobile robots are a different category. You can learn more about general facility upgrades in our overview of warehouse automation systems.

Amazon's own page linked below is the place to check current details on its robots.

Can an Independent Warehouse Buy Kiva Robots Today

Independent warehouse operators cannot purchase Kiva robots or Amazon Robotics hardware. Amazon manufactures these drive units exclusively for its own fulfillment centers and does not sell the machines to third-party facilities.

Facilities seeking equivalent material handling capabilities must source from commercial vendors that build equipment within related automation categories. You can review available market options in our directory of commercial warehouse robots for sale.

Buyers evaluating alternatives should divide the market into four distinct purchasing categories depending on the specific workflow requirement:

• Commercial goods-to-person systems that move proprietary shelf pods or totes to stationary pick stations.
• Autonomous mobile robots (AMRs) that transport pallets, carts, or loose containers across open facility layouts.
• Robotic arms equipped with machine vision to pick individual items from bins, as explored in our guide to warehouse picking robots.
• Warehouse execution software (WES) that manages order queues, inventory allocation, and fleet traffic.

Selecting the Right Fulfillment Architecture

Deploying mobile robotics requires aligning the mechanical hardware with your facility throughput goals and physical building constraints. Confirm with each vendor what floor, layout, and shelving a system needs, since requirements differ between pod-carrying systems and AMRs.

This type of mobile automation is not the right choice for operations that handle irregular bulky freight, lumber, or odd-sized industrial machinery that cannot fit onto standardized shelving pods. Facilities with low daily order counts or slow-moving stock generally achieve better returns by optimizing manual picking paths rather than funding mobile robot hardware.

Our perspective on adopting alternative mobile systems would shift if an operation handled high-mix, rapid-turn small parcel inventory that required hundreds of daily worker steps per hour. In those high-density environments, moving the storage medium to the worker remains the most direct way to reduce travel overhead.

Where Advanced Automation Is Heading Next

Humanoid robots are one prospective path for warehouses, as covered in our analysis of humanoid robots in warehouses.

Regardless of future form factors, Kiva Systems proved that coordinating autonomous vehicles through centralized software could dramatically compress fulfillment cycle times. The principles of automated dispatch, fleet collision avoidance, and goods-to-person delivery continue to govern commercial warehouse robotics today.