Figure AI robots exceed 24 hours of warehouse operation test
A fleet of humanoid robots developed by Figure AI has sustained continuous warehouse sorting operations for more than 24 hours, marking a major milestone in autonomous industrial robotics. The demonstration began as a planned eight hour live test on 13 May and evolved into an open ended endurance run after the system completed its initial goal without errors.
The robots operated in a logistics environment where they sorted parcels on a conveyor belt, scanned barcodes, and directed packages without human remote control. Three units alternated tasks to maintain continuous workflow, handling soft envelopes and boxed shipments at steady speed. By the 17 hour mark, the system had processed more than 22,000 parcels, showing consistent performance under extended operational pressure. The run continued beyond 24 hours, and later reports indicated totals exceeding 47,000 parcels after 38 hours of uninterrupted work.
At the core of the system is Helix-02, a unified neural network designed to merge vision, touch, and proprioception into a single control architecture. The model coordinates walking, balance, and manipulation simultaneously rather than separating movement and handling into distinct modules. This integrated design allows the robots to respond continuously to real world input while maintaining stability and precision in object handling tasks.
The operation also relied on multi robot coordination and automated fault management. When a unit detects a malfunction, it performs self diagnostics and moves independently to a maintenance zone while another robot replaces it in the workflow. This structure reduces downtime and removes the need for human intervention in routine recovery processes. Reported sorting speed averaged around three seconds per parcel, placing performance close to human warehouse workers in similar conditions.
The company has also accelerated its industrial rollout. Production at its BotQ facility has shifted from one robot per day to one per hour within a few months, with more than 350 humanoid units already assembled. The long term target includes scaling annual output to 12,000 robots and reaching 100,000 units within four years, signaling an aggressive push toward large scale deployment in logistics environments.
-
17:00
-
16:45
-
16:30
-
16:27
-
16:15
-
16:08
-
16:00
-
15:52
-
15:47
-
15:45
-
15:30
-
15:25
-
15:17
-
15:15
-
15:00
-
14:59
-
14:45
-
14:40
-
14:30
-
14:22
-
14:15
-
14:10
-
14:00
-
13:45
-
13:42
-
13:33
-
13:30
-
13:15
-
13:00
-
12:45
-
12:30
-
12:15
-
12:00
-
11:53
-
11:45
-
11:30
-
11:20
-
11:15
-
11:04
-
11:00
-
10:45
-
10:43
-
10:35
-
10:35
-
10:30
-
10:25
-
10:20
-
10:17
-
10:15
-
10:12
-
10:00
-
09:46
-
09:45
-
09:44
-
09:39
-
09:30
-
09:18
-
09:15
-
09:04
-
09:00
-
08:51
-
08:45
-
08:40
-
08:30
-
08:23
-
08:15
-
08:07
-
08:00
-
07:51
-
07:45
-
07:37
-
07:30
-
07:18
-
07:15
-
07:03
-
07:00
