Micro-fulfillment is one of my favorite topics to write about. Because of the research I’ve conducted on the subject and my hands-on experience with robotics and automation, I’m frequently referred to as an expert in micro-fulfillment and robotics. The purpose of this article is to educate and inform.
Every year at GroceryShop, automation vendors tell some version of the same story: robots, speed and a lower cost to serve. Most audiences nod along without asking the question that matters. It isn’t simply whether a system works. It is who built it, why it was built and whether it has been proved in the operating environment where it will be deployed.
That question is about to be tested on the industry’s highest-profile stage. Walmart, the largest grocery retailer in the United States, has placed its bet on a vendor that didn’t design the store-level system it is now developing.
I want to examine that decision because it is the clearest live case study of a question every retailer attending GroceryShop will eventually face: When automating stores, should a retailer choose proven, purpose-built grocery technology or an industrial automation company’s attempt to adapt a system it acquired?
Walmart and Symbotic’s Unique Relationship
Walmart’s automation story starts with Alert Innovation, an independent robotics company that designed and built the original micro-fulfillment system now at the center of Walmart’s plan.
Walmart could have partnered with AutoStore early in the process. Based on my conversations with individuals involved in that decision, I did not believe the evaluation adequately accounted for AutoStore’s capabilities or grocery-specific operating advantages. Based on my experience, I believe AutoStore would have been a better choice for Walmart than Alert Innovation.
H-E-B, considered by many to be the best grocery retailer in the United States, conducted a rigorous evaluation. The company assessed the available micro-fulfillment systems, including those from Takeoff Technologies, Fabric, Alert Innovation, Dematic and AutoStore. H-E-B chose AutoStore.
Full disclosure: I assisted with the selection process. As a former adviser to Takeoff and Fabric, and someone with extensive knowledge of Alert Innovation and AutoStore, I believed AutoStore was the most logical choice. H-E-B has since installed approximately a dozen AutoStore systems across its network.
Now that Ocado has introduced its Store Based Automation solution, I encourage H-E-B to pilot the technology and determine whether it should maintain its current AutoStore strategy, partner with Ocado for SBA or operate a hybrid model using both platforms.
What most coverage of Walmart’s choice has missed is where Alert Innovation’s technology originated.
Alert Innovation founder John Lert had been trying to automate grocery fulfillment since 1994. In 2007, Lert founded CasePick Systems with funding from Rick Cohen, the owner of C&S Wholesale Grocers, to build a mobile robotic shuttle system for warehouse automation. Cohen acquired full ownership of the company in 2009 and renamed it Symbotic. Cohen is Symbotic’s founder, chairman and CEO today.
Lert left in 2011, still convinced that the larger opportunity was a true three-dimensional shuttle system rather than the system Cohen had taken over. He founded Alert Innovation in 2013 to build it.
Lert frequently contacted me and asked for my opinion. During one discussion about AutoStore and his Alphabot system, I asked him a direct question: Was Alphabot better than AutoStore?
Lert replied, “No, it’s not. AutoStore is a better system.”
I have tremendous respect for John Lert. His honesty was refreshing, and we were able to discuss how his platform could eventually become better than AutoStore.
Things changed rapidly for Alert Innovation in 2019, when Walmart contracted the company to install an Alphabot system at a store in Salem, New Hampshire. Additional installations followed, and Walmart acquired Alert Innovation in 2022.
In January 2025, Symbotic agreed to acquire Walmart’s Advanced Systems and Robotics business for $200 million in cash, with additional contingent consideration possible. Walmart also agreed to fund a $520 million development program under which Symbotic would create an automated pickup-and-delivery system for Walmart stores. The platform is now branded SymMicro.
Symbotic’s core expertise, demonstrated through its work with Walmart, Albertsons, C&S Wholesale Grocers and Target, is large-scale distribution-center automation: cases and palletized freight moving through massive facilities. Symbotic is a leader in warehouse automation.
SymMicro represents the company’s first attempt to apply that expertise to store-level grocery fulfillment using an underlying system Symbotic did not originally design and only recently acquired.
I’ve used this comparison before, and it remains appropriate: It is the equivalent of Ford deciding to build a new muscle car by taking an F-150 pickup truck and modifying the drivetrain instead of purpose-designing a car from the frame up. The F-150 is an excellent truck. It was never meant to be a Mustang.
On Symbotic’s August 2026 earnings call, Rick Cohen discussed the staged development and validation process for SymMicro. The first installation is an overbuilt prototype. A redesigned version is expected to reduce costs, require less space and refine the system’s functionality.
According to Cohen, Walmart is not expected to authorize the full 400-store deployment until the redesigned prototype has been validated, with that decision currently anticipated around early 2028.
That is not caution for its own sake. It indicates that Symbotic is still validating whether the redesigned platform can satisfy Walmart’s cost, footprint and operating requirements at scale.
Now compare that timeline with the alternative Walmart passed on.
Ocado Is the Leader in Grocery Automation
Ocado took the opposite path from the beginning. It spent 25 years operating as an online grocer and building automation to solve its own operating challenges: low margins, high SKU variability, multiple temperature zones, inventory accuracy and complex fulfillment requirements.
Only after building technology for its own grocery operation did Ocado become a technology provider to other retailers.
The scale is worth putting directly in front of this audience because it is easy to underestimate from outside the category.
By the end of the first half of 2026, Ocado’s On-Grid Robotic Pick technology had generated more than 350 million picks to date and had been deployed across 14 customer fulfillment centers operated by eight partners. Ocado also reported 27% year-over-year growth in international CFC volumes.
Individual sites can operate fleets of more than 1,000 robots, process millions of items each week at accuracy rates exceeding 99.9% and operate robotic pick stations capable of handling more than 600 items per hour.
The newest hardware, the 600 Series robot, is approximately five times lighter than its predecessor and is manufactured primarily through additive manufacturing. Its lower weight allows Ocado to install lighter grids more quickly and economically inside a wider range of buildings.
Ocado’s newest product, Store Based Automation, or SBA, is a compact application of the company’s established technology. It is designed to fit inside or beside an existing store rather than requiring a standalone, 350,000-square-foot customer fulfillment center.
SBA is not a new company’s first attempt at grocery automation. It is a 25-year grocery automation leader’s answer to the problem retailers are trying to solve: how to automate an existing footprint without committing to a nine-figure standalone facility.
None of this means Ocado has been beyond criticism.
I have criticized Ocado in several articles. I publicly called for the company to build a micro-fulfillment system in 2020, 2022 and early 2025. I argued that it was inexcusable for Ocado to remain focused almost exclusively on large customer fulfillment centers while the market moved toward smaller, store-based formats.
Ocado formally announced its intention to introduce Store Based Automation in November 2025.
What About Kroger?
Anyone who has followed Ocado’s U.S. story is already thinking about Kroger. The issue deserves a direct answer rather than a dodge.
Kroger closed three of its Ocado-powered customer fulfillment centers in early 2026 and canceled another planned facility. Kroger paid Ocado $350 million in compensation. That is real, and it happened.
What it was not, in my opinion, was primarily a technology failure. It was a site-selection, network-design and demand-forecasting failure. The affected facilities were built in markets that did not generate the order density required to make a 350,000-square-foot automated CFC economical.
This is a point I can address directly. I was the person who recommended that Kroger acquire or partner with Ocado, and I designed a different strategy for deploying Ocado’s technology from the one Kroger ultimately implemented. Had Kroger followed my recommendations, far fewer CFCs would have been constructed.
The same underlying Ocado platform that struggled to achieve the required economics at several Kroger locations has produced a different result for Coles in Australia, where the facilities were placed in markets with sufficient demand.
Coles reported that its two automated CFCs achieved positive EBITDA during their second year of operation, supported by higher volumes and increased on-grid robotic picking.
The technology is substantially the same. The market conditions and implementation decisions are different.
That is the real lesson. It is not a verdict that SBA should replace large CFCs everywhere.
A large-format CFC still makes sense where order density supports it. That is why Coles continues to operate its two Australian CFCs and why Kroger is not abandoning large CFCs altogether. Kroger’s restructuring retained the model in its highest-density markets.
The failure was not the CFC concept in every market. The failure was building capital-intensive facilities in locations that could not generate enough volume to justify them.
Kroger’s emerging hybrid approach is the more accurate model for GroceryShop attendees to consider: retain large CFCs where demand density supports them while using capital-light, store-based automation in markets where a standalone facility cannot be justified.
Match the fulfillment format to the market. Do not select one automation architecture and force every location into it.
Lotte Is the Case to Watch
Lotte Shopping in South Korea is the case worth watching as this strategy develops globally.
Ocado signed its partnership with Lotte in November 2022. Construction of the first CFC in Busan began in December 2023, and the facility went live in August 2026. The Ocado-powered operation serves a catchment covering approximately four million households across Busan and the broader Yeongnam region.
The facility introduced several Ocado innovations, including a fully automated freezer operating at temperatures as low as minus 25 degrees Celsius, On-Grid Robotic Pick and support for South Korea’s “dawn delivery” standard, which places fresh orders on customers’ doorsteps before 7 a.m.
Lotte and Ocado plan to build as many as six CFCs across South Korea by 2030, targeting five trillion won in online grocery sales. A second facility is planned for Goyang to serve the Seoul metropolitan area.
As Ocado’s fulfillment strategy evolves to combine large CFCs with compact Store Based Automation, depending on market density, Lotte’s remaining rollout is a natural candidate for the same hybrid model: full CFCs in the densest regions and SBA in markets that do not justify standalone facilities.
Lotte has not publicly confirmed that SBA will be part of its specific rollout, but the opportunity is clear.
The Question Every Retailer Should Be Asking
This is not merely a Walmart story. It is a category story.
Every retailer walking the floor at GroceryShop will face some version of the same decision during the next 24 months: whether to automate stores and which vendor’s technology to choose.
The instinct in grocery is often to default to the vendor with the existing relationship, the familiar name or the lowest perceived switching cost. Walmart’s SymMicro decision is a live example of that instinct: choosing continuity with a known vendor instead of publicly demonstrating a rigorous comparison among the leading grocery-automation platforms.
I recommend that every retailer evaluating store-level automation conduct the comparison Walmart has not yet run publicly: a genuine side-by-side pilot between purpose-built grocery automation platforms and adapted industrial systems under identical conditions before committing capital at scale.
Ocado and AutoStore should both be included in that evaluation.
I have said publicly that I believe Walmart CEO John Furner will ultimately choose Ocado’s SBA over SymMicro once a direct comparison takes place. Whether or not that prediction proves correct, the underlying discipline remains valid: pilot before committing, and weigh the vendor’s operating record as heavily as its sales presentation.
Twenty-five years of grocery-specific engineering is not a marketing claim. It is a track record that can be tested.
Make every vendor prove its claims before signing a contract that may determine your automation strategy for the next decade.
Based on multiple conversations with industry sources, I believe Walmart, H-E-B, Ahold Delhaize, Kroger, Amazon, Albertsons and Target have either evaluated, discussed or are expected to discuss Ocado’s SBA offering.
These conversations may be preliminary. I am not suggesting that any agreement, pilot or deployment is imminent. I am saying that leading retailers are examining the technology—and every grocery executive evaluating store-level automation should do the same.