Inside Project Kobe: Amazon's plan to build Walmart-style supercenters powered by warehouse robots and AI
Amazon Challenges Walmart with Ambitious New Supercenter Format
Amazon is escalating its competition with Walmart through a bold new retail concept that merges traditional supercenter shopping with advanced warehouse automation, according to internal company documents reviewed by Business Insider.
The initiative, codenamed "Project Kobe," represents one of Amazon's most significant bets on transforming brick-and-mortar retail. The format blends grocery and general merchandise with an integrated automated fulfillment center, enabling in-store shopping, curbside pickup, and delivery operations from a single location.
Expansion Plans Taking Shape
While Amazon previously disclosed plans for an Orland Park, Illinois, location near Chicago, the documents reveal additional sites advancing in Cherry Hill and Edison, New Jersey, plus Oak Brook, Illinois. If initial locations prove successful, the company envisions potentially rolling out dozens more stores.
Each facility spans approximately 225,000 square feet—comparable to Walmart Supercenters in size but fundamentally different in operation. Roughly 100,000 square feet, or nearly half the building, is allocated to automated warehousing, far exceeding the 18,000 to 36,000 square feet typically dedicated to fulfillment at Walmart stores, according to Cornell University supply chain professor Linwei Xin.
"To my knowledge, there is no clear precedent," Xin said.
The stores are designed to stock approximately 250,000 products, nearly twice the assortment of a typical Walmart location.
Significant Cost Challenges
The ambitious format comes with substantial tradeoffs. Internal projections indicate fulfillment costs per item will run about 12% higher than Amazon's existing sub-same-day delivery network and significantly more than its most automated facilities.
Grocery items present particular challenges, with costs estimated at nearly $2 per unit—approximately 10% above sub-same-day delivery expenses. A major factor: many fresh items must be retrieved from the sales floor, requiring workers to walk over 500 feet to bring products back to the fulfillment area.
The Orland Park site alone is projected to require $33 million in capital investment.
As Amazon scales to higher-volume markets where grocery demand could triple, these inefficiencies may intensify. Future locations are expected to move more grocery inventory to the warehouse section to ease sales floor congestion, though this creates new challenges around refrigeration capacity and storage space.
Robotics and Automation at the Core
The warehouse section relies on AutoStore's robotic storage system, which uses a grid of stacked bins with robots traversing the top to retrieve items. Amazon chose the system partly for its storage density, maximizing inventory within limited space.
However, Amazon is also developing its own proprietary system called "Orbital," designed to handle ambient, chilled, and frozen products in a single compact automated setup. The technology remains roughly two years from deployment and could feature in future Kobe locations.
Despite heavy automation, significant manual labor remains necessary. Bulky items require a two-level picking system, while fresh groceries are often selected directly from store shelves. Individual orders may pull from multiple locations throughout the facility, all requiring consolidation within tight timeframes.
AI-Driven Merchandising
Behind the scenes, Amazon is building artificial intelligence capabilities to optimize product selection for each location. The company is developing tools that allow category managers to set strategic objectives while AI models generate recommendations based on demand patterns and space limitations.
The system, powered by internal optimization models and a custom AI assistant named "Frida," aims to minimize manual planning and enable scalable decision-making across multiple stores. Documents indicate Amazon intends to transition from human-curated selections to fully automated, model-driven assortments.
However, the technology remains under development. One document acknowledged the company has not yet reached its "North Star," underscoring how much remains unproven.
Competitive Context
Amazon's push into large-format retail comes as it continues trailing Walmart significantly in grocery. Despite surpassing Walmart in total revenue last year, Amazon commands just 3% of the U.S. grocery market compared to Walmart's 21%, according to Numerator data. Walmart's extensive supercenter network also enables same-day delivery access to 93% of American households.
An Amazon spokesperson disputed these figures, arguing that Numerator's methodology undercounts Amazon's grocery presence by excluding non-food consumer packaged goods like toothpaste, diapers, and shampoo. When including non-perishables, the spokesperson said, Amazon ranks among the nation's largest grocers with over $150 billion in gross sales in 2025.
Project Kobe operates under the leadership of Jason Buechel, who serves as both Amazon's Vice President of Worldwide Grocery and CEO of Whole Foods. Since assuming the role last year, Buechel has prioritized restructuring Amazon's grocery operations and tightening integration with Whole Foods.
The initiative remains highly confidential within Amazon, with approximately 50 employees across Seattle, New York, and Austin directly involved. Planning documents are tightly controlled.
Early Stage Development
An Amazon spokesperson characterized the supercenter concept as being in "early development," noting that details will continue evolving.
"The new supercenter concept we are testing will make it convenient for customers to shop a broad selection of quality products at great value across fresh groceries, household essentials, and general merchandise—all in one trip," the spokesperson said.
The first Kobe location in Orland Park is scheduled to open in late 2027, providing Amazon's first real-world test of whether this ambitious hybrid model can succeed where previous physical retail experiments have struggled.