← All briefings
Warehouse roboticsFleet schedulingPilot evaluation

China Warehouse Robots: Can Every Robot Charge Before Dispatch?

Four robots need 30 minutes each before the same dispatch wave. Draw the dock bookings to see why three chargers can miss the deadline despite enough total minutes.

China AI Tour Research3 min read
Conceptual warehouse scene with a worker scanning a parcel beside a loaded trolley and an open delivery vehicle.
Conceptual warehouse scene with a worker scanning a parcel beside a loaded trolley and an open delivery vehicle.

Four mobile robots must be ready for a dispatch wave at 13:00. All four become available to charge at 12:20. Each needs 30 uninterrupted minutes in a compatible dock. How many docks would you ask the warehouse operator to demonstrate? Under those assumptions, four. A spreadsheet that divides 120 docking minutes by the 40-minute break suggests three, but nobody can give the fourth robot a complete slot.

Draw one line per dock

Use the time when the robot can actually enter the dock as the start of its window. Use the latest time it can leave the dock and still reach its next assignment as the end. In this hypothetical exercise, travel time is zero, each dock serves one robot at a time and the 30 minutes include docking and release. The duration is an assumed service requirement for the exercise.

With three docks, start robots A, B and C at 12:20. They leave at 12:50. Robot D then has ten minutes before dispatch. The three docks together have 30 unused minutes during those last ten minutes, but the robot can occupy only one dock at a time. Its earliest finish is 13:20. Changing which robot goes last changes its identity, not the missed deadline.

For identical uninterrupted sessions sharing one release time and deadline, a dock can finish only the whole number of sessions that fit in the window. Here that is one per dock. Four docks can serve all four robots from 12:20 to 12:50. This establishes slot availability under the assumptions. The operator still needs to show that the power supply supports the simultaneous load and that each robot actually reaches its required charge.

Test a change in timing before pricing another dock

Keep three docks and the same 13:00 deadline. Suppose dispatch can release A at 12:00 while B, C and D still become available at 12:20. Try this schedule:

DockFirst bookingSecond booking
1A, 12:00–12:30D, 12:30–13:00
2B, 12:20–12:50Available after 12:50
3C, 12:20–12:50Available after 12:50

This arrangement finishes all four sessions. It borrows 20 minutes of A's availability from the preceding work period and gives D no margin at the deadline. Ask which picks A would have made during those 20 minutes. Then move A's arrival five minutes later and replay this booking plan. D would finish at 13:05. The dispatcher can now compare a concrete service risk with the cost of a fourth dock.

Replay the receiving site's dispatch wave

For a China warehouse robotics pilot, request a redacted log with robot ID, earliest arrival, compatible dock IDs, actual plug-in and release times, charge on departure and next assignment deadline. Include waits for a blocked aisle or occupied dock. A controller that reports only active charging time can omit the minutes that determine whether a robot reaches dispatch.

Use observed durations for the battery conditions and target charge in the intended workload. If charging can be interrupted, or several docks share a power limit that changes their speed, rebuild the schedule using those operating rules. A feasible drawing on paper is a reason to run the trial. Record which robots returned on time, which assignments waited and which earlier tasks were displaced to make the plan work.

MEASURE 2.3 of NIST's AI Risk Management Framework 1.0 (2023) calls for demonstrating performance under conditions similar to deployment. Use an actual dispatch wave to test the fleet controller's booking decisions. The logistics evaluation guide follows the affected orders through loading and handover.

From the Wire

News behind the questions

Read China AI Wire ↗

Robotics ·

Tiangong robots integrate Agora technology for remote operation ↗

Tiangong robots have integrated Agora's real-time interaction technology for remote operation, Beijing Business Today reported on October 25. The Beijing Humanoid Robot Innovation Center is also advancing Tiangong 3.0 applications across several practical settings.