Global Logistics Facility Tech Weekly

Global Warehouse & Logistics Tech Digest

Period covered 12 – 18 October 2026 Written by Tomo Miyata

This week I chose five stories about the software that instructs material handling equipment and people, rather than about how well the equipment performs.

My lead story is a report from IoT Analytics, a German market research firm. The market for the equipment and software that move goods inside logistics facilities, store them and control their flow was worth $96 billion in 2025. The firm forecasts about $133 billion by 2030. Software was $6.3 billion of that in 2025, and it is growing at 11.4% a year, the fastest of the six major segments. Software here includes the warehouse management system (WMS), which manages inventory and shipping; the warehouse control system (WCS), which tells material handling equipment what to move next and where; and the warehouse execution system (WES), which decides in what order and how incoming orders are turned into tasks.

The same report says that software able to instruct material handling equipment from any maker is appearing between the WMS and the automation equipment. The background is that companies running facilities do not want to redo the connection to their WMS every time they add robots. DHL, the international logistics group, used software from SVT Robotics, a U.S. maker of robot-integration software, to connect new robots to its facility systems in seven countries. The work took a few hours and needed no SVT engineers on site. The report also lists a trend: AI agents, meaning AI that carries out tasks on its own, are starting to take over the decisions facility managers make about staffing and restocking.

In the United Kingdom there is a facility that runs on instruction software its operator built itself. THG Fulfil, the logistics-services arm of the British group THG Ingenuity, set up a facility for the sneaker retailer Footasylum that uses 85 AutoStore robots. AutoStore is an automated storage system in which robots run over stacked containers and lift them out; the company of the same name supplies it. Before construction, THG Fulfil ran 80,000 simulations to decide the number of robots and workstations. The facility went fully live on 14 September. System uptime is 99.8%, and it uses fewer than half the picking workstations of Footasylum's previous operation while handling 212% of the previous volume. The WCS that instructs the facility's material handling equipment is one THG Fulfil developed in-house over five years.

Amazon shows what happens when you add more material handling equipment. Its robots now number more than one million. Many robots running through narrow aisles cause congestion, and Amazon deals with it using DeepFleet, a set of AI models that predict congestion in advance. There is also research on the effect on employees. A study published in an academic journal in 2025, using records from Amazon facilities, reported that facilities with robots had 40% fewer serious injuries than conventional facilities but 77% more minor injuries from repetitive motion. A co-author, a professor, sees the robots setting the pace of work as one cause of the minor injuries.

I expect robots themselves to keep getting easier to obtain, including through robots-as-a-service (RaaS), a contract in which you rent robots as a service instead of buying them. This week AutoStore announced robotic picking on a RaaS basis for small businesses. IoT Analytics also says in its report that Chinese makers are entering Europe and North America at lower prices. Once the robots themselves are easy to get, what will separate facilities is whether one piece of software can give order and destination to material handling equipment from different makers and to people alike. A second difference is whether the company running the facility has drawn a line on how much to leave to AI. Take a delayed truck: is it the AI or a person who changes the staffing? Without that line, if a problem occurs after AI is introduced, neither the operating company nor the people on the floor will know who is responsible.

If you add material handling equipment without a system to instruct it, goods wait in front of the equipment and the equipment stops more often. Amazon having to build a separate dedicated model to predict congestion is one example. GROUND's product, GWES, is built around software that instructs the material handling equipment and the people in a logistics facility. It is not a configuration in which other vendors' software was combined afterwards. Japanese logistics facilities face a severe labour shortage and a flow of many kinds of goods mixed together. I believe that whether a facility has a mechanism to instruct its material handling equipment and its people will decide the difference in productivity.

GROUND Inc. — Founder & CEO Tomo Miyata

This Week's News

Five stories

01 Lead story

The interior of a large logistics facility with automated equipment — Equipment inside a logistics facility (illustrative). Photo: Unsplash

Market researchWork instruction softwareAI agents

IoT Analytics — Facility logistics is a $96 billion market in 2025, and software is the fastest-growing major segment at 11.4% a year

IoT Analytics, a German market research firm, published the contents of its "Intralogistics Market and Technology Trends Report 2026" on October 15. Intralogistics means moving, storing and controlling the flow of goods inside a logistics facility or factory site. The market covers equipment, software, communications infrastructure and services. It was worth $96 billion in 2025 and has grown 8% a year since 2020. The firm forecasts 6.7% annual growth through 2030, to about $133 billion. The report covers 31 product and service categories and draws on more than 30 interviews with equipment makers, system integrators and operating companies.

The fastest-growing major segment is software, rising from $6.3 billion in 2025 to $10.8 billion in 2030 (11.4% a year). It includes the WMS, which manages inventory and shipping; the WCS, which instructs material handling equipment; the WES, which decides how orders are assigned to work; digital-twin software, which recreates real equipment in a virtual space so it can be tested; and analytics software. Spending on autonomous mobile robots and automated guided vehicles (AMRs and AGVs), which belong to the equipment segment, nearly doubles from $4.3 billion in 2025 to $8.4 billion in 2030, about 14% a year. The analysts note that Chinese makers are improving performance, building local service networks and lowering prices. The top challenge cited by companies running facilities is labour shortage; according to the report, annual staff turnover reaches 200% to 300% in facilities with a lot of picking work.

Two of the three trends the firm names involve software. The first is that AI agents, meaning AI that carries out tasks on its own, are starting to make decisions that facility managers have made, such as staffing, restocking and maintenance planning. EPG of Germany, a maker of software for logistics facilities, presented software at the LogiMAT 2026 trade show in March in which AI agents change staff assignments and trigger restocking, and which acts when a truck is delayed. The second is that software able to instruct material handling equipment from any maker is appearing between the WMS and the automation equipment. DHL used software from SVT Robotics, a U.S. robot-integration software company, to connect new robots to its facility systems in seven countries within hours, without SVT engineers. One of the report's conclusions is that operating companies are starting to use robots from several makers. As a result, VDA 5050, a common communication standard drawn up by the German Association of the Automotive Industry (VDA) and others, and software that works with any maker will become criteria when companies buy.

Source: IoT Analytics / 2026.10.15

An operator checking an automated conveyor system in a logistics facility — Photo: Unsplash

Case studyWCSUnited Kingdom

THG Fulfil — Footasylum's automated facility goes live, handling 212% of its previous volume with fewer than half the workstations

THG Fulfil, the logistics-services arm of Britain's THG Ingenuity, has brought into full operation an automated facility for Footasylum, a sneaker and streetwear retailer. It is the first time THG Fulfil has installed AutoStore (an automated storage system in which robots run over stacked containers) for an outside retailer. The facility was announced in June, processed its first order on August 28 and went fully live on September 14. It is the first peak season Footasylum will go through on this system.

The facility has 85 AutoStore robots and 96,000 container slots, and holds Footasylum's entire range on sale. According to THG Fulfil, system uptime is 99.8%, throughput is 212% of Footasylum's previous operation, and it uses fewer than half the previous number of picking workstations. Before construction, THG Fulfil and Footasylum ran 80,000 simulations using Footasylum's order history, seasonal demand and product data to decide the number of robots and workstations.

The facility's material handling equipment is controlled by a WCS (the software that instructs material handling equipment) that THG Fulfil developed in-house. It spent five years developing it while running it in several AutoStore facilities it operates itself. Even after going live, it says, it will keep improving the order in which containers are retrieved, how orders are grouped and where stock is placed, using real order data.

Source: FashionNetwork / 2026.10.15

Monitoring consoles in a control room lit in blue — Photo: Unsplash

Robot operationsCongestion forecastingImpact on workers

Amazon — More than one million robots, congestion predicted by AI, and a study finding 77% more minor injuries at robotic facilities than at conventional ones

Ars Technica, a U.S. online publication, reported on October 12 on how Amazon runs its robots. Amazon operates more than one million robots and employs more than 1.5 million people (according to its 2025 annual report). Most of the robots are wheeled mobile robots that carry shelving units to workers. Amazon says Sequoia, the system that coordinates the movements of many robots, speeds up identifying incoming inventory and storing it on shelves by up to 75%. When many robots run through narrow aisles, congestion occurs. Predicting it is the job of a separate set of AI models called DeepFleet.

The growth is in robotic arms such as Robin, which sorts packages; Cardinal, which loads packages into carts; Sparrow, which repacks the contents of containers; and Vulcan, which has tactile sensors. Vulcan has been running in facilities in Spokane in the U.S. and Hamburg in Germany since 2025, and is due to be installed at more sites in 2027. The arms number in the thousands, and the chief financial officer said on the July earnings call that the company intends to more than double them this year.

On the effect on workers, there is a study published in the journal ILR Review in April 2025. Using Amazon records from 2016 to 2019, it found that robotic facilities had 40% fewer serious injuries and 77% more minor injuries from repetitive motion than conventional facilities. Co-author Brad Greenwood, a professor at George Mason University, says the robots setting the pace of work is one cause of the minor injuries. Amazon counters that since 2019, worldwide, its recordable incident rate has fallen 43% and its rate of incidents with lost time has fallen 70%. The article also reports that robotic facilities need about 30% more staff in maintenance, reliability and technical roles than conventional ones.

Source: Ars Technica / 2026.10.12

Cardboard boxes moving along a conveyor — Photo: Unsplash

New productsRaaSPredictive maintenance

AutoStore — Robotic picking for small businesses and predictive maintenance in its fall 2026 product announcement

AutoStore, the Norwegian maker of automated storage systems, announced its fall 2026 products and features on October 13. It points out that even automated facilities still rely on manual handling for unexpected situations in areas such as handling items of different sizes, picking by case and by piece, maintenance, inventory checks and receiving. It says this announcement is aimed at reducing that manual work.

In the fourth quarter of 2026 it will offer FlexBinsPort, which handles containers of different heights at one workstation, and expects storage efficiency to rise 10 to 15%. AutoCase, due in the first quarter of 2027, handles cases up to the industry-standard 600 by 400 mm, can process the same batch of work in parallel at several workstations, and already runs at Nowaste Logistics in Sweden and Komplett in Norway. Pio Pal, for small businesses, offers robotic picking on a RaaS basis, a contract to rent it as a service instead of buying it. In packing it can process up to 540 order lines an hour, and combined with pre-picking it can raise throughput up to three times without adding workers. The Norwegian clothing brand Famme already uses it.

CubeFleet, a new predictive-maintenance feature, uses operating data and maintenance thresholds to warn in advance when maintenance will be needed and even creates the work order. AutoStore expects it to cut unplanned stoppages by 35 to 45%. It also announced CubeExplore, which uses generative AI to estimate return on investment, and CubeStudio, which uses AI to assist with design. The company has more than 1,300 customers in 68 countries and about 2,000 installed systems.

Source: Business Wire (AutoStore announcement) / 2026.10.13

A factory production line with industrial material handling equipment — Photo: Unsplash

AcquisitionSortation equipmentGermany

Symbotic — Acquires Germany's cellumation, gaining sortation equipment that decides a package's destination as it moves it

Symbotic of the U.S., which builds AI-driven robotic systems for logistics (and supplies automation equipment to Walmart's U.S. logistics facilities), announced on October 13 that it will acquire cellumation, a company in Bremen, Germany. The price was not disclosed. cellumation developed the sortation equipment celluveyor. It consists of hundreds of cells, small conveying modules that can move packages in any direction, each controlled independently. The equipment does not just carry packages; it decides their destination as it moves them.

Symbotic expects to raise throughput and handle a wider range of items in less floor space than conventional conveyor sortation. It says the layout of inbound and outbound robot cells can be simplified, freed-up floor space can go to storage, and equipment costs may fall. celluveyor is used in facilities run by postal, e-commerce and logistics operators.

cellumation keeps its name and continues selling to existing customers, and will also sell to system integrators other than Symbotic. Its founders, Hendrik Thamer (chief executive) and Claudio Uriarte (chief technology officer), will continue to run it. Symbotic Chairman and CEO Rick Cohen said this follows its recent acquisitions as the purchase of a company with a standalone product, and that cellumation's equipment will be sold to direct customers and to other integrators and built into Symbotic's systems.

Source: Robotics & Automation News / 2026.10.13