Period covered 26 October – 1 November 2026Written by Tomo Miyata
This week I chose five stories on investment in, and deployment of, logistics automation. The lens I used is which software takes charge of instructing material handling equipment (conveyors, sorters, robots and similar) and people.
The lead story is a demonstration published by Intel of the United States and SAP of Germany. A work request issued in SAP's business software is carried out by robots in a logistics facility. Intel is a leading chipmaker, and SAP a leading maker of enterprise business software. The SAP Extended Warehouse Management (EWM) system used in the demonstration is a warehouse management system (WMS), software that manages inventory and work in a facility. When a request entered in EWM reaches the robots, they carry materials and report the status and completion of the task back to EWM. Everything from request to report is connected automatically.
A logistics facility often runs, alongside the WMS, software that assigns work and instructs material handling equipment. A warehouse execution system (WES) decides in what order incoming orders are broken into tasks, and a warehouse control system (WCS) tells the equipment inside the facility which unit to move next and where. In this demonstration, it has not been disclosed which software assigns a request arriving from EWM to which robot, and in what order. As results, Intel cited shorter transport distances and less repetitive manual handling, but gave no figures.
There was also a large investment announcement. Walmart, the US retail giant, said on October 29 that it will expand its partnership with Symbotic, a US company that builds robot-and-AI automation systems, with an initial commitment of about $6 billion. The programme covers 25 regional distribution centers in the United States. Symbotic will deliver everything from storage to shipping as one system, and Walmart chose to hand that scope to a single company. Berkshire Grey, a US company owned by SoftBank Group, added AI to the software that runs its own robots, to adjust the flow of goods and staffing as demand changes. What I am watching is less the size of the investment than which company's software issues instructions to the whole facility.
Growth in logistics automation also appears in earnings. KION of Germany, a world leader in forklifts and logistics automation, reported that in the first nine months its automation division's revenue rose 16.9% and its adjusted operating profit rose 36.6%. The forklift and related division, however, took fewer orders than expected, and KION lowered its 2026 outlook. Adoption of humanoid robots is still small. GXO, the US logistics group, is adding Digit, a two-legged robot from Agility Robotics of the United States, at two more US sites. Patrick Kelleher, GXO's chief executive, said on an earnings call that none of the roughly 20,000 robots GXO is deploying in 2026 will be a humanoid in full production. He also said the robots have not yet delivered a return that justifies the cost. Those remarks are as reported by Martin Cid Magazine.
I expect the robots themselves, and the AI models that drive them, to keep getting easier to obtain and cheaper. If so, two points will separate productive facilities from the rest. The first is whether one piece of software can instruct material handling equipment from different makers, and people, on the order of work and where to go. The second is whether the operating company can draw a line on how much the AI is allowed to decide. For example, Berkshire Grey's unloading robot detects items that need a person's attention. Should such an item go to a person at once, or be held until a natural break in the work? If the rule is not set in advance, the item has no destination and work stops.
Adding material handling equipment without software that instructs it and people creates problems. Each maker's software cannot read the status of another maker's equipment, so goods pile up in the handoff from one unit to the next, and stoppages become more frequent. A single-system approach covering storage to shipping, as at Walmart, avoids this more easily, in my view. But few companies can spend about $6 billion and rebuild facilities over several years. I think Intel and SAP built their demonstration to work with any robot maker because facilities that use equipment from several makers need exactly that.
GROUND designs GWES as software that instructs material handling equipment and people together, not as other vendors' software combined afterwards. Japan's logistics facilities face a severe labor shortage and handle a wide variety of goods in a single site. That is why I believe whether a facility has software that instructs material handling equipment and people shows up as a difference in productivity.
GROUND Inc. — Founder & CEOTomo Miyata
This Week's News
Five stories
01Lead storyEquipment inside a logistics facility (illustrative). Photo: Unsplash
Business software integrationPhysical AIUnited States / Germany
Intel and SAP — Robots carry out requests entered in a logistics facility management system and report back, in a joint demonstration
Intel, the US chipmaker, said on October 28 that it had worked with SAP, the German business software company, on a demonstration in which robots in a logistics facility carry out requests entered in business software. SAP Extended Warehouse Management (EWM) is software that manages inventory and work in a logistics facility. In the demonstration, a work request entered in EWM reaches autonomous robots. The robots carry Intel hardware and software for edge AI, meaning AI that runs on the equipment at the site. They move materials and report task status and completion back to EWM.
SAP handles the business information on which operation a task belongs to, the approval steps and the work rules. Intel provides the technology that lets a robot perceive its surroundings, choose a route and move. According to the two companies, transport distances became shorter and repetitive manual handling decreased. They gave no figures. The demonstration was shown live at SAP TechEd, SAP's technology conference.
The design is meant to work regardless of robot maker, operating system, AI model or deployment environment (cloud or on-premises). Intel supplies a computing base that combines CPUs, GPUs and NPUs, along with development software such as OpenVINO. John Healy, Intel's head of industrial and robotics, said the demonstration answers customers' wish to connect AI directly to physical operations. Dr. Lukasz Ostrowski, who leads physical AI (AI that drives real-world equipment) and robotics at SAP, said the design lets customers build AI-driven automation into core operations while keeping the freedom to choose their robot platform. The two companies say they intend to turn the design into a standard that can be extended across sites, robots and uses.
Major investmentSingle-vendor automationUnited States
Walmart — Expands partnership with Symbotic to automate 25 US regional distribution centers, with an initial commitment of about $6 billion
Walmart, the US retail giant, said on October 29 that it will expand its partnership with Symbotic, a US maker of AI-driven robotic automation systems. The initial commitment is about $6 billion. Walmart will add automation over several years to 25 regional distribution centers in the United States, the large sites that gather and sort goods before they go out to stores. Symbotic will also support further automation across Walmart's global logistics network.
In Symbotic's system, AI software decides where each case is stored, and fast mobile robots store and retrieve it. Retrieved cases are stacked on pallets and shipped. Walmart says the equipment doubles the number of cases a distribution center handles per hour compared with a conventional one. By the end of 2025, about 65% of Walmart's stores were receiving goods from distribution centers with this level of automation.
Separately from this partnership, Walmart has already committed to buy and install systems for 400 automated sites at its stores, used for order pickup and delivery. Rob Montgomery, senior vice president for Walmart US supply chain operations, said the partnership should improve service, add capacity, simplify associates' work and cut costs. Rick Cohen, Symbotic's chairman and chief executive, said the relationship began more than ten years ago with a single product and has grown into an end-to-end offering across Walmart's supply chain.
Berkshire Grey — Adds AI that adjusts the flow of goods and staffing to BG Pilot, the software that runs its robots
Berkshire Grey, a US maker of robots for unloading and picking, said on October 29 that it had strengthened the AI in its robot product line. The company is a wholly owned subsidiary of SoftBank Group. The upgrades are based on its AI software, BG Pilot, which draws on records of billions of robotic picks. The company collected those records over ten years of operations in customer sites. Its customers include FedEx, the US logistics group, Meijer, a US supermarket chain, and Fanatics, a US sports merchandise retailer.
The upgrades reach three products. In Scoop, a trailer-unloading robot, an AI model processes camera and 3D data together to detect the trailer wall, obstructions and items that need a person's attention. In Stride, used for store replenishment, AI adjusts the flow of goods, assignments to workstations and staffing as demand, shipping priorities and operating constraints change. In Core, used for packing, AI chooses the box, the placement of items and a tight packing pattern, taking into account protection and handling requirements.
The company says new business opportunities have risen by more than 65% over the past six months. Dave Paratore, chief executive, said demand for automation in logistics facilities and interest in physical AI are leading to more deals, and that BG Pilot will be the foundation for meeting that demand. The company says BG Pilot can cope with a wide variety of goods and packaging, unexpected events and changes in staffing. These statements come from the company's announcement.
KION — Automation division revenue up 16.9% and adjusted operating profit up 36.6% in the first nine months; outlook cut on weak forklift division
KION of Germany, a world leader in forklifts and logistics automation that also owns the automation systems maker Dematic, reported results for the first nine months of 2026 on October 29. Group revenue rose 3.9% to €8.522 billion and adjusted operating profit rose 9.9% to €632.5 million, a margin of 7.4% (7.0% a year earlier).
The automation division, Intelligent Automation Solutions, grew revenue 16.9% to €2.519 billion and adjusted operating profit 36.6% to €172.2 million, lifting its margin to 6.8% (5.8% a year earlier). Orders were €2.575 billion, 12.4% below a year earlier, when large project orders in the second quarter reached a record. Orders in July to September were €750.8 million, 1.4% above the same period a year earlier.
The forklift division, Industrial Trucks & Services, saw revenue fall 0.7% to €6.034 billion, and its orders did not grow as expected. KION says the division did not keep pace with a growing market, particularly in low-priced vehicles. On October 14 KION cut its 2026 outlook. It lowered group revenue guidance from €11.525–12.025 billion to €11.450–11.750 billion, and adjusted operating profit guidance from €880–980 million to €815–875 million. It narrowed the automation division's guidance within the earlier range: revenue of €3.350–3.450 billion and adjusted operating profit of €230–250 million. KION now expects to reach its medium-term target of a double-digit operating margin only after 2027.
GXO — Adds Agility Robotics' two-legged Digit robot at two more US sites, while saying the robots have not yet paid for themselves
GXO, the US logistics group, and Agility Robotics, the US maker of the two-legged humanoid robot Digit, said on October 26 that they will add Digit at two more US sites, in Statesville, North Carolina, and Lebanon, Tennessee. Digit will take on repetitive material-moving work in e-commerce fulfillment. Deployment starts in early 2027 with the current Digit 4, and will switch to Digit 5, which Agility unveiled in September, once it is commercially available.
At the existing site in Flowery Branch, Georgia, Digit works alongside existing automation, placing totes (containers for goods) on conveyors. It has moved more than 100,000 totes so far. GXO rents the robots under a contract signed in June 2024, in a robots-as-a-service (RaaS) arrangement, rather than buying them. The companies describe it as the first formal commercial deployment of humanoid robots. Agility's cloud platform, Arc, handles robot management and the link to the facility's existing operations. Digit has logged more than 65,000 hours across all Agility customers.
Neither company disclosed the cost of the two sites or when they expect to break even. Patrick Kelleher, GXO's chief executive, said on the latest earnings call that GXO has run 45 humanoid trials, according to Martin Cid Magazine. He added that the robots have not yet delivered a return that justifies the cost and that this is a couple of years away. None of the roughly 20,000 robots GXO is deploying across its sites in 2026 will be a humanoid in full production, he said. Agility plans to list through a merger with Churchill Capital Corp XI, a special purpose acquisition company (SPAC), expected to close in the fourth quarter of 2026.
