Global Logistics Facility Tech Weekly

Global Warehouse & Logistics Tech Digest

Period covered 2 – 8 November 2026 Written by Tomo Miyata

I believe that adding machines or AI alone will not raise a logistics facility's productivity. This week, vendors overseas stood out for building instructions to machines and people, and limits on what AI may decide, into their software. The lead story is an essay from EY.

The lead story is an essay published on November 7 by Abhi Manohar of EY, the accounting and consulting giant. It starts from an observation: logistics facilities are far more automated than before, yet facility managers still spend much of their time reacting to changing conditions. One example is a picking robot running at full speed while orders cannot be completed because replenishment stock has not arrived. Another is a conveyor moving cartons quickly while a change in a carrier's pickup time forces supervisors to reprioritize outbound shipments by hand.

Two kinds of software issue work instructions in a logistics facility. A warehouse execution system (WES) decides in what order incoming orders are broken down into tasks and how they are assigned. A warehouse control system (WCS) tells the machines in the facility which one should move what, and where, next. EY argues that faster machines do not raise a facility's overall throughput if decisions are slow when conditions change. It recommends a process in which a system spots an order at risk and proposes a response, a person approves it, and the system follows through on execution.

Where a system uses AI, how much to let AI decide is a choice for the company that runs the facility. One way to split it: AI alone reorders replenishment, while a person approves moving staff who are mid-task to another station. This week, two vendors built that dividing line into their products. Oracle, the US enterprise software giant, says that for its new AI feature that carries out work, people set goals and policies and give the important approvals, while the AI acts within its permissions and keeps a record of what it does. Blue Yonder, the US supply chain software company, launched an order management product. In it, AI spots orders at risk of delay and lays out options, and a person decides within the rules and permissions set by the company that deploys it.

I think AI capabilities are becoming easier to adopt because they come built into existing supply chain software. Oracle put a screen that shows priorities for facility work into the quarterly update it released on November 6. Blue Yonder's new product can be adopted starting with its core functions, without replacing surrounding systems.

That leaves two things that I think will separate one facility's productivity from another's. The first is whether a single piece of software can tell machines from different makers, multiple systems and people what to do in what order and where. The second is whether people can draw the line on how much AI is allowed to decide.

Adding machines or AI without software that issues those instructions leaves work backing up between machines and between systems, and the whole facility stopping more often. In the example EY gives, a picking robot cannot complete orders at any speed if replenishment does not arrive. In a joint survey by the research firm Everest Group and Lean Solutions Group, 69% of organizations had moved AI beyond isolated uses, but only 19% had substantially redesigned workflows for collaboration between people and AI. The original article does not say whom the survey covered. GXO, the US contract logistics giant, is bringing the multiple labor management systems it ended up with, partly through acquisitions, onto one common platform.

GROUND's product, GWES, is built around software that instructs both machines from different makers and people. We did not design it as a bolt-on to other vendors' software. Japan's logistics facilities face severe labor shortages and handle a wide mix of goods in a single facility. For that reason I believe that having software that instructs both machines and people will determine the productivity gap between facilities.

GROUND Inc. — Founder & CEO Tomo Miyata

This Week's News

Five stories

01 Lead story

Illustration of a logistics facility — Equipment inside a logistics facility (illustrative). Photo: Unsplash

EssayInstructing machinesUnited States

EY — Facilities are heavily automated, but coordination between machines is still done by hand, an essay argues

Abhi Manohar of EY (Ernst & Young), the accounting and consulting giant, contributed an article on logistics facility automation to the US trade outlet Supply & Demand Chain Executive on November 7. Robots, conveyors, automated storage and retrieval systems (AS/RS) and warehouse management systems (WMS, software that manages inventory and work in a facility) have spread, the article says, yet facility managers still spend a great deal of time reacting to changing conditions. It cites the 2025 report of the US industry group MHI, in which 45% of the supply chain executives surveyed said they intend to adopt automated solutions.

The article names three challenges. First, receiving, putaway, replenishment, picking and shipping are each managed with their own automation and metrics, and compete for labor, equipment and storage space. Second, exceptions such as stock shortages, equipment downtime and last-minute order changes cannot be resolved by fixed procedures. Third, dashboards show that shipments are running late, but there is no means of deciding the best response and getting it done in time.

Improvement is described as a progression in three stages. The first is to automate so that fixed tasks run reliably. The second is to use forecasting and optimization analytics to see congestion and delays coming. The third is to decide, approve and follow through on responses across the facility's processes. As an example, the article takes a case where a production delay holds up replenishment and an order due to ship that day is at risk. The system identifies the order at risk, compares options such as moving replenishment up or reassigning labor, and proposes a response. Once a person approves, the system monitors execution and checks whether the risk of delay has gone.

Existing equipment need not be replaced, the article says, and it lists four points for proceeding. The first is to start with exceptions that recur. The second is to make WMS records, inventory, order priorities and equipment status available as the same information. The third is to decide which actions the system may carry out on its own and which need human approval. The fourth is to measure exception resolution time, on-time shipment, supervisors' workload and cost, and to extend to other workflows only after results are confirmed.

Source Supply & Demand Chain Executive (contributed by EY) / 2026.11.07

Order managementAI agentsUnited States

Blue Yonder — AI-built order management software spots orders at risk of delay and lays out options

Blue Yonder, the US maker of supply chain planning and execution software, on November 6 launched order management software built around AI, for retailers and wholesale distributors. It is a new product in its Cognitive Solutions portfolio. It links the stages of an order, from the delivery promise through fulfilment, customer service and returns, as one flow, and connects commerce, planning and execution systems around each order, so that sourcing, inventory allocation and responses can be revisited when conditions change. Gurdip Singh, Blue Yonder's chief product officer, says every order touches almost every part of the supply chain.

At the centre are AI Fulfilment and Customer Service Agents. They identify orders at risk, bring the relevant data together on one screen and explain the trade-offs of possible actions. A person on staff then decides, within the rules and permissions the deploying company has set.

According to the company, the software draws on real-time inventory across stores, distribution centres, suppliers, inbound shipments and returns that can be resold, so it can give accurate delivery promises. When supply, demand or capacity changes, it flags the affected orders and proposes reallocating them and re-promising delivery dates in bulk. For returns, it applies business rules and profitability criteria to decide whether goods are resold, repaired, sent back to the vendor or routed elsewhere. The product is cloud-native and can be adopted starting with core functions, without replacing surrounding systems. The company says its order management systems processed 7 million order reservations on Black Friday 2025 and more than 133 million estimated-delivery-date calculations. It serves more than 3,000 retailers, manufacturers and logistics providers.

Source eCommerceNews Australia / 2026.11.06

Operator checking equipment inside a logistics facility — Photo: Unsplash

AI agentsSupply chain softwareUnited States

Oracle — Fusion Claw, a new feature that carries AI-driven work through to completion, ships in the 26D update of its supply chain software

Oracle, the US enterprise software giant, on November 6 released the 26D update of Oracle Fusion Cloud SCM, its cloud supply chain management software. The main new feature is Fusion Claw. AI-driven business applications analyze a problem, plan a response and carry it through to completion within defined business rules and permissions. In one example, a product team asks whether a bill of materials is ready for launch; Fusion Claw gathers the relevant information, evaluates options against business constraints and recommends next steps. It carries out actions within the rules and monitors progress as conditions change.

People set goals and policies and give the important approvals. Fusion Applications, Oracle's suite of business software, enforces permissions and business rules and keeps an auditable record of actions. In 26D, Fusion Claw powers the working screens for product launch readiness, sourcing and supply planning allocation. This is according to a blog post by Daniel Bachar, Oracle's vice president of supply chain product management.

For logistics facilities, the Warehouse Execution Command Center shows priorities across fulfilment activities and helps teams deal with bottlenecks and use labor and equipment more effectively. In inventory, AI agents recommend packing configurations and alternative storage locations, reducing handling and manual lookups. For quality, Oracle added screens that detect emerging inspection risks and process instability.

Source Oracle blog / 2026.11.06

Cardboard boxes moving on a conveyor — Photo: Unsplash

AI platformContract logisticsUnited States

GXO — AI platform GXO IQ to reach more than 50 sites this year; labor management systems to be unified

GXO, the world's leading contract logistics company, based in the US, reported results for July–September 2026. Revenue rose 5.7% from a year earlier to $3.6 billion, and growth excluding acquisitions was 6.4%. New business wins were $375 million (up 34%), and the sales pipeline reached a record $3.1 billion. Chief executive Patrick Kelleher answered questions from the US trade publication Logistics Management.

Kelleher said GXO aims to bring GXO IQ, its AI platform for logistics facilities, into more than 50 sites this year. GXO IQ has two roles: middleware for deploying the company's AI tools, and a data lake that holds the data the AI uses. The company is developing AI agents in-house and with outside partners. In one example, GXO demonstrated for a fashion retail customer a tool that signals when replenishment is needed. It does more than show that a picking location is about to run out of stock; using past ordering patterns, it signals when to act before a stockout.

GXO is also moving the multiple labor management systems it has, partly as a result of its acquisitions, onto one common platform. It expects to have deployed it by the end of 2027 at sites covering about 60% to 70% of total labor spend. The company says it will make it easier to move people between sites and to spread improvements made at one site to others, and will widen the room to add AI to labor management. Kelleher also said buying robots or AI is not enough; the company needs experienced teams who can combine people, technology and processes.

Source Logistics Management / 2026.11.05

Illustration of logistics — Photo: Unsplash

SurveyAI adoptionUnited States

Everest Group and Lean Solutions Group — 69% of organizations have taken AI beyond isolated uses, but 19% have substantially redesigned workflows

According to a report published jointly by the research firm Everest Group and Lean Solutions Group, 69% of organizations have moved AI beyond isolated use cases. Only 19% have significantly or extensively redesigned workflows so that human experts and AI can work together. The share that has redesigned workforce roles to a comparable degree is 15%. The article reporting the survey does not say whom it covered or how many responded.

Jack Freker, chief executive of Lean Solutions Group, says technology can analyze, recommend and automate at scale, but people remain accountable for the outcome. The organizations that get the most from AI, he says, will combine human experts, AI technology and redesigned workflows.

Source Supply & Demand Chain Executive / 2026.11.03