JD Logistics Unveils Five New “Wolf” Robots Across the Supply Chain

The strategy is an interesting counterpoint to the humanoid boom. JD Logistics is not trying to solve the entire supply chain with one general-purpose machine. Instead, it is building specialized robots for specific workflows and coordinating them through the Super Brain 3.0 orchestration layer. The result looks more like an operating system for physical logistics than a single robot launch.

Five new Wolf Pack elements at JDD 2026

Material attributed to JD Logistics' official channel and an independent on-site report from National Business Daily confirm five new products or technologies. Coverage also highlights an upgraded Agile/Alien Wolf dexterous arm that is already being deployed at scale in sorting, so RoboMorrow does not count that as a sixth completely new product.

Warehouse Wolf: mobile picking inside an existing warehouse

The most relevant product for conventional fulfillment is Warehouse Wolf (仓狼), a mobile embodied-AI picking robot. JD says it can be deployed without rebuilding the whole facility and claims an installation cycle of roughly two weeks. Manufacturer figures include 99.9% picking accuracy, more than 85% whole-warehouse SKU coverage and about 80 UPPH. These numbers come from JD Logistics; RoboMorrow has no independent benchmark comparing them with human picking, AMR-assisted workflows or competing mobile manipulators.

The deployment model is still important. If a robot can enter existing aisles without a six-month automation rebuild, it reduces one of the biggest barriers to adoption. For European operators, integration cost and downtime can matter more than the sticker price of the robot itself.

Smart Wolf Low-Temperature: goods-to-person at -20°C

The second element is a low-temperature Smart Wolf system for cold storage. JD describes operation at -20°C using WOS global control, active defrosting and contactless power. The company claims 100% higher storage capacity, 200% higher operating efficiency and a 10% reduction in unit cost after deployment.

Those figures depend heavily on the baseline warehouse, so RoboMorrow does not treat them as universal benchmarks. Cold storage is nevertheless a difficult automation environment: batteries, lubrication, condensation and sensors behave differently from room-temperature facilities. If the system performs reliably in live operations, that is a more meaningful proof point than another controlled humanoid demonstration.

Mother Wolf: a 24/7 automated pharmacy

Mother Wolf Health targets pharmacy and retail fulfillment. JD describes end-to-end automation from medicine receiving to dispensing, mixed batch and SKU handling, 24/7 unattended operation, 99.9% dispensing accuracy and up to 90 orders per hour. The company also claims full compliance traceability.

In pharmaceutical automation, throughput is only part of the problem. Batch tracking, expiry dates, exception handling and accountability for a wrong dispense matter just as much. The most useful future evidence will therefore include certification and human-oversight procedures, not only order rate.

Lone Wolf sixth-generation Plus: L4 delivery capability

The new Lone Wolf is a sixth-generation Plus delivery vehicle with claimed L4 autonomous-driving capability. For JD, that links the warehouse and last-mile layer without moving planning into a completely separate operating system. L4 remains a capability label from the vendor; legal road operation depends on regulators, operating domain and deployment conditions.

Flying Wolf L05: logistics drone for unmanned operations

Flying Wolf L05 is a logistics drone that JD describes as supporting fully unmanned operation. It extends Wolf Pack beyond the warehouse and road network. The relevant commercial questions are not only range and payload, but aviation permissions, landing infrastructure, weather envelope, safety and cost per mission.

Super Brain 3.0 may be the most important product without wheels or legs

The Super Brain layer is intended to coordinate robots across storage, sorting, transport and delivery. That orchestration layer may be JD's biggest strategic advantage. At large scale, the problem is no longer whether one robot can pick a parcel; it is how thousands of different machines share tasks, energy, queues, routes and exceptions.

If the architecture works as described, system value increases with the number of robot types. Warehouse Wolf can prepare an order, Smart Wolf can manage totes, sorting systems can route the parcel onward, Lone Wolf can cover road delivery and Flying Wolf can handle selected aerial missions, all under a shared planning layer. That is a more mature physical-AI story than a single impressive hardware demo.

Why we are not recycling the “3 million robots” headline

JD again references a five-year plan involving 3 million robots, 1 million autonomous vehicles and 100,000 drones. That target had already been publicly discussed before 9 September, so RoboMorrow does not treat it as a fresh standalone story. Today's event is the full Wolf Pack presentation and the concrete new products and technologies shown at JDD.

This is a core deduplication rule: a new conference and a new secondary article do not turn an older strategy target into a new event. The volume plan remains useful context for scale, not the primary hook.

What is a JD claim and what remains unknown

• JD claims: 99.9% picking accuracy, 80 UPPH and 85% SKU coverage for Warehouse Wolf; cold-store efficiency gains; 90 orders/hour for Mother Wolf; and L4 capability for Lone Wolf.
• Unknown: list prices for most systems, TCO, integration cost and standard service terms.
• No independent benchmark yet covers uptime, intervention rate or cost per operation across the new family.
• European availability is unclear for the JDD 2026 configurations.

Why Poland and Europe should care

For European logistics operators, JD shows an alternative route to physical automation: not one humanoid for everything, but a portfolio of specialized machines under one orchestration system. That model may be easier to justify economically in a high-volume warehouse where tasks are well defined and measurable.

Poland's logistics market is large enough for comparable architectures, but real competitiveness will depend on local service, WMS integration, spare parts, CE compliance and deployment cost. Performance inside JD's Chinese network does not automatically establish ROI in a European distribution center.

RoboMorrow decision

PUBLISH as a new PL/EN article. We found no existing RoboMorrow owner for this specific JDD 2026 Wolf Pack launch. The next alert should wait for an overseas rollout, price, major external customer or independent operating metrics.

Sources

JD Logistics Blackboard reposted by Sina — full Super Brain + Wolf Pack description · National Business Daily — on-site JDD 2026 report · JD Logistics 2025 Annual Report — context on earlier robot deployments