A Shenzhen resident named Li signed for a learning tablet on December 1, 2025, and accidentally became a statistic. His package was the 180-billionth parcel shipped in China that year, and almost no human hand moved it. Warehouse robots sorted it, an autonomous van carried it to his neighborhood, and a courier handled only the last doorstep meters. The machine behind the milestone was JD Logistics' "wolf" fleet — and it is already at work.
Key takeaways
- JD Logistics reported 2025 revenue of about RMB 217.1 billion (≈ US$30.2 billion / HK$235 billion).
- Its Shenzhen "Zhilang" (智狼) warehouse runs nearly 200 handling and lift robots in a goods-to-person system.
- The "Dulang" (独狼) sixth-generation delivery vehicle operates at L4 autonomy across more than 20 provinces.
- JD says it plans to buy 3 million robots, 1 million unmanned vehicles, and 100,000 drones over five years.
- The 180-billionth 2025 parcel moved through this stack with no human touch in storage, sorting, or last-mile transfer.
The "wolf" system in the warehouse
Inside JD's Shenzhen small-items warehouse, nearly 200 "Zhilang" handling robots and lift robots work with high-density shelves and custom totes. The company says a single order can go from received to exited in as little as 15 minutes, and that the system lifted inbound and putaway efficiency by roughly six times versus manual handling. Pick accuracy is quoted at 99.99 percent.
This is not a research pilot. JD says "Zhilang" goods-to-person sites operate in nearly 20 cities, with a first overseas warehouse opening in the UK in the fourth quarter of 2025. The appeal is density: the same floor space stores more stock and retrieves it faster, which matters when a retailer holds millions of SKUs.
The van that meets the courier
At the neighborhood end, JD's "Dulang" sixth-generation autonomous vehicle runs prescribed routes from a delivery station to a community gate, where a courier takes over for the doorstep. JD describes it as L4-capable and says thousands of such vehicles already operate across more than 20 provinces. The courier keeps the relationship and the hard last meters; the robot removes the dead time of driving between stops.
The 180-billionth parcel is the cleanest proof: a customer order triggered the warehouse robots, an autonomous van did the line-haul to the gate, and the courier completed delivery. JD calls the pairing a "human-vehicle CP," and couriers on record say it raised their pickup-and-delivery volume.
The five-year bet
JD Logistics' loudest number is a procurement plan: 3 million robots, 1 million unmanned vehicles, and 100,000 drones across its supply chain within five years. Even allowing for ambition inflation, the direction is clear — JD is spending on embodied AI (具身智能) as a cost lever, not a demo. Its "SuperBrain 2.0" (超脑2.0) model acts as a planner, simulating a billion-parcel network in under three minutes and adjusting routes in real time. For a firm of JD's scale, even partial success rewrites the unit economics of Chinese e-commerce logistics.
Beyond China's borders
JD's robot stack is not staying domestic. The company said its first overseas "Zhilang" warehouse opened in the UK in late 2025, and its drones and autonomous vehicles have been tested in Saudi Arabia, including a first overseas drone flight there in 2025. For a logistics operator, exporting the "SuperBrain plus wolf fleet" model is a way to monetize software it already built for its own network.
The labor question follows. JD frames robots as augmenting couriers, not replacing them, and it has created new frontline roles — "smart warehouse maintainers," "unmanned-vehicle operators," and "drone pilots." Whether those roles scale as fast as the robots is the open debate. What is clear is that JD is treating robotics as core infrastructure, not a side experiment, and measuring success in parcels moved per human hour.
Honest limitations
Warehouse robots and fixed-route vans are the easy part of automation. The doorstep — stairs, gates, odd addresses, and human interaction — still needs a person, and JD's model keeps the courier central rather than removing them. The 3-million-robot pledge is a forward statement, not a booked order, and past logistics automation promises have slipped. Drone delivery, while tested on mountain and island routes, remains a small share of volume and faces airspace and safety limits. JD's 2025 profit still relied on network density, not robot labor savings alone.
What readers can do now
- If you run a fulfillment operation, benchmark JD's goods-to-person model: ~200 robots, 15-minute order-to-exit, 99.99 percent pick accuracy are concrete targets to ask vendors about.
- If you ship with JD or a rival, expect more fixed-route autonomous vans at the station-to-gate leg; plan courier workflows around robot drop-offs.
- If you invest in automation, watch component makers — lithium cells, L4 compute, and warehouse robot joints — since JD's scale is where those suppliers win or lose.