The encyclopedia · Engineering & Operations · Operational decision · 2017–2021
JD.com ran its robot warehouses on dispatching algorithms, not more pickers
JD.com put operations research at the core of its unmanned warehouses: algorithms dispatch AGVs to racks, lifting efficiency 5x and cutting storage cost 50%.
JD.com · JD Logistics
the move
JD.com built its own couriers and network because China's parcel volume exploded nearly 30-fold in a decade, and humans who had to walk every aisle could not keep up.
Its answer was the unmanned warehouse: automated guided vehicles carry racks to pickers rather than sending pickers to racks. That only pays off if the robots are dispatched well, so operations research controls which robot does which job, how racks are pooled, and how time is used.
The result was warehouse operating efficiency up to five times, average storage cost down by half, and hundreds of millions of dollars in savings, with 5,000 robots working across 100 of JD's 800 buildings. The project reached the INFORMS Franz Edelman Award finals in 2021.
why it works
- The walking was the cost: carrying a rack to a human replaces a long walk with a short lift and a return trip.
- Real-time dispatching keeps robots busy instead of idle, so each machine earns its capital cost.
- Automation eased a genuine labor crunch while letting volume scale without a matching headcount.
- A faster, cheaper path to the same order also improved speed and customer experience.
what transfers
When the bottleneck is movement, optimizing how orders reach the work is cheaper than adding more people to walk further.
what came after
JD scaled the system to 5,000 robots in 100 warehouses and said it would grow its robot fleet toward 100,000 over five years. The work was named a 2021 INFORMS Edelman Award finalist, and the dispatching research behind it was later published by the operations-management group that collaborated with the company.
references
spotted an error? The archive wants to know.