GROUND GROUND
Warehouse Tech Weekly
VOL. 30 2026.11.22

Global Logistics Facility Tech Weekly

Global Warehouse & Logistics Tech Digest

Period covered 16 – 22 November 2026 Written by Tomo Miyata

Connecting material handling equipment from different makers requires something beyond installing the hardware.

This week's first story is the distribution center of C&S Wholesale Grocers, the large US grocery wholesaler, which Symbotic opened to the press. Inside, 155 Symbotic robots travel through an eight-level automated storage structure, storing and retrieving goods case by case. Even though Symbotic builds its own storage structures and robots, it chose other makers' robotic arms for depalletizing incoming cases and other makers' lifts for carrying cases to each level.

Symbotic's software is used to decide the robots' routes and the order in which cases are retrieved, while controllers from Rockwell Automation, the US industrial controls giant, are used to control the whole installation, including the other makers' arms and lifts. In terms of the system hierarchy, the roles of a warehouse execution system (WES), which breaks shipping orders down into tasks and sets their sequence and assignment, and a warehouse control system (WCS), which controls the material handling equipment, are divided between Symbotic's software and Rockwell's controllers.

This week also brought two announced acquisitions of robot makers by large controls and automation companies. Rockwell agreed to acquire Agilox, an Austrian maker of autonomous mobile robots (AMRs), and Witron of Germany acquired SOLWR Robotics of Norway. Witron says SOLWR's case-picking robot can be used not only with Witron equipment but also on its own, connected to other vendors' WMS. Even as large companies keep buying robot makers, the material handling equipment installed in a single logistics facility will not necessarily all come from one company.

The difficulty of connecting material handling equipment from different makers is discussed in detail in a report by Plattform Lernende Systeme, a German industry-academia body. According to the report, connecting communications and unifying data formats are not enough; robots' actions can be coordinated only once systems share the meaning of the information exchanged and the effect that actions based on it will have on work on the floor. Applied to shipping, even if an autonomous mobile robot sends a signal that it has arrived at a shelf, the next task cannot be assigned unless the receiving system knows which shipping order's items are to be picked from that shelf.

Integration with existing systems is also an issue for AI. In a survey of about 2,000 people by a lab at the Massachusetts Institute of Technology (MIT) and Mecalux of Spain, 40% of respondents named integration with existing systems as the biggest obstacle to expanding the use of generative AI. In the same survey, 51% of AI agent use cases decide and execute inventory movements without human approval, compared with 22% for commitments to customers and 12% for financial matters. For now, the decisions entrusted to AI are mainly inventory movements within the facility.

Even Symbotic, which builds its own storage structures and robots, relies on software and controllers separate from its hardware to direct work across an installation that includes other makers' equipment. At GROUND, we have designed GWES so that it does not depend on any one maker's hardware and assumes it will direct work for each maker's material handling equipment in a facility, as well as for the people on the floor. In Japan's logistics facilities, which handle many product lines and where workers and material handling equipment share the same floor, I believe a system that can direct work for both, regardless of maker, is all the more necessary.

GROUND Inc. — Founder & CEO Tomo Miyata

This Week's News

Five stories
01 Lead story
A logistics facility with conveyors and automated equipment
Equipment inside a logistics facility (illustrative). Photo: Unsplash

DeploymentEquipment controlUnited States

Symbotic — Opens a C&S grocery distribution center to the press, integrating 155 of its robots with other makers' arms and lifts through Rockwell controllers to build store-specific pallets automatically

Symbotic, the US maker of automation systems for logistics facilities, opened to the press a distribution center that C&S Wholesale Grocers, the large US grocery wholesaler, operates in Connecticut, the US trade outlet Control Global reported on November 19. The tour coincided with Automation Fair 2026, the trade show of US industrial controls giant Rockwell Automation. Symbotic's customers include Albertsons, Walmart, Target and Medline. Symbotic's backlog stands at $22.5 billion.

After workers scan incoming pallets and remove the wrapping, robotic arms depalletize the cases layer by layer. The cases are separated one by one, and scanning records their dimensions, weight, lot numbers, expiry dates and other details. Cases with damaged packaging are diverted to a separate lane for a person to check. Each inbound cell handles about 1,215 cases an hour. Other facilities built by Symbotic reach up to 1,700.

Lifts carry the cases to each level of the eight-level storage structure, where 155 Symbotic robots, called SymBots, store and retrieve them. Symbotic's software decides how the robots move, where inventory is placed and in what order cases are retrieved for orders. Storage is free-location: open space is assigned to match each case's size, and fast-moving products are placed near the outbound docks.

The robots and storage structure are Symbotic's, while the robotic arms and lifts come from other makers. Rockwell's controllers, drives and networking equipment control the whole installation, including the other makers' equipment and safety systems. On the outbound side, cases are resequenced by destination store and built into pallets that follow each store's layout. Products that sit in the same store aisle go on the same pallet, reducing the work of stocking shelves at the store. Placement is decided by weight, size and crushability, and the outbound side handles up to 1,350 cases an hour. Building one pallet takes four to six minutes. "The bots are ours, the storage structures ours, and then we integrate everything together with the help of Rockwell Automation," said David Clear, Symbotic's vice president of sales.

Source Control Global / 2026.11.19

Reviewing tabulated data on a laptop
Photo: Unsplash
02

SurveyAI agentsGlobal

MIT and Mecalux — 57% expect AI to make more than a quarter of facility operating decisions by 2028; fewer than 40% of use cases have quantitatively confirmed significant gains

A lab at the Massachusetts Institute of Technology (MIT) Center for Transportation & Logistics and Mecalux, the Spanish storage systems and software company, published a survey on the use of AI in logistics facilities. About 2,000 supply chain and logistics facility professionals in 21 countries responded. 57% expect AI to make more than a quarter of their facility operating decisions autonomously by 2028, and 90% said they will increase AI investment over the next two to three years.

Among the use cases respondents described in detail, 87% of generative AI and 81% of AI agent applications improved their target metric at least moderately. But only 39% of generative AI and 36% of AI agent use cases have quantitatively confirmed significant gains. 40% of companies have stopped at least one AI project in the past two years. Use cases with multiple AI agents working together account for 11%.

Among AI agent use cases, those that decide without human approval account for 51% for inventory movements, 42% for data corrections, 22% for customer commitments and 12% for financial decisions. The most common form of oversight has AI acting within set limits while people monitor and step in when needed. Companies that have set operating rules for AI were more likely to confirm significant gains quantitatively: 51% versus 26% for generative AI and 46% versus 23% for AI agents. 40% named integration with existing systems as the biggest obstacle to scaling generative AI, and 54% said better integration with WMS and enterprise resource planning (ERP) systems would do most to accelerate adoption.

Source Modern Materials Handling / 2026.11.18

An industrial robotic arm in a factory
Photo: Unsplash
03

AcquisitionAutonomous mobile robotsUS / Austria

Rockwell Automation — Agrees to acquire Austrian autonomous mobile robot (AMR) maker Agilox from Carlyle

Rockwell Automation, the US industrial controls giant, has agreed to acquire Agilox of Austria from the US investment firm Carlyle, the German trade journal Technische Logistik reported on November 18. The deal is expected to close in the first half of 2027. The price was not disclosed.

Agilox develops and builds autonomous mobile robots that carry parts and goods in factories and logistics facilities. Its robots are designed to reduce manual transport and fixed conveyors. The company has its own technology for AI-based route decisions and for multiple robots adjusting to each other's positions as they move. It serves more than 400 customers in automotive, consumer goods, pharmaceuticals, industrial manufacturing and logistics, with more than 40 sales and deployment partners in over 35 countries.

Source Technische Logistik / 2026.11.18

Cases stacked on a pallet with automated equipment
Photo: Unsplash
04

AcquisitionPicking robotsGermany / Norway

Witron — Acquires Norway's SOLWR Robotics, adding a case-picking robot that can be retrofitted into existing logistics facilities

Witron, the German company that designs and builds automation systems for logistics facilities, has taken over SOLWR Robotics of Norway, the German trade journal Technische Logistik reported on November 19. SOLWR was founded in 2016 and has about 50 employees. It has two main products. One is "Grab," a robot that drives through aisles on its own and stacks cases of up to 30 kg onto pallets for stores. The other is "Sort," a unit that sorts empty pallets by quality and type, handling 300 pallets an hour as standard and 600 or more when expanded.

Grab can be used in ambient and fresh areas. A version for frozen areas is planned. It can operate in areas where workers are also present. Witron says Grab can be combined with Witron systems or connected on its own to other vendors' WMS, with little integration effort on both the mechanical and IT sides. Investment costs and integration times were not given. The main target is existing logistics facilities that want to automate without rebuilding, including small and mid-sized retail facilities. SOLWR Software AS is excluded from the acquisition.

Source Technische Logistik / 2026.11.19

A production line with multiple robotic arms
Photo: Unsplash
05

ReportInteroperabilityGermany

Plattform Lernende Systeme — Report sets out four levels for connecting robots from different makers, and what large AI models can and cannot do

Plattform Lernende Systeme, a German body through which experts from industry and academia make recommendations on AI, has published a report, the German trade journal Technische Logistik reported on November 17. Its subject is interoperability: how robots from different makers can work together. According to the report, differing interface specifications and control software that is not open to outsiders get in the way when different robots are combined. Until now this has been handled by standardizing interfaces, protocols and data formats, which takes effort and lacks flexibility.

The report divides interoperability into four levels: connecting systems to each other, aligning data formats, aligning the meaning of the information exchanged, and understanding what actions based on that information will cause on site. Only at the last level, it says, can robots coordinate their actions well.

The report says large language models, and AI models that handle language, images and actions together, can help at the level of aligning meaning and beyond. It also lists limitations: a lack of on-site activity data and industry-specific data, incorrect outputs, a limited grasp of cause and effect, and an inability to fully translate objects' positions and actions into precise instructions. As a result, interoperability can temporarily break down and create safety risks. In the journal's article, Johannes Hinckeldeyn, who leads advanced technologies at Germany's KION Group, said users, robot makers and researchers need to work together to bring these ideas into practical use.

Source Technische Logistik / 2026.11.17

← Back to Warehouse Tech Weekly Photos: Unsplash  © 2026 GROUND Inc.