X Square's WALL-B sorts 10,000 packages in 5 hours, outpacing Figure by ~53%

At WRC 2026, X Square demonstrated its WALL-B robot with a self-developed robotic arm in a 5-hour-14-minute continuous logistics sorting challenge: 10,000 packages handled, averaging 1.88s per item, a throughput of 1,911 packages per hour — beating its previously published record of 1,816/hour (98%+ accuracy) — while showing strong long-run stability.

Confirmed

• Challenge lasted 5h14m with 10,000 packages processed, 1.88s each, 1,911 per hour
• WALL-B follows an embodied-AI "one brain, many bodies/tasks" approach and uses X Square's own robotic arm
• Multiple posters, including Scobleizer, ChrisGPT and rohanpaulai, relayed the same live test video and figures

Why it matters

• ChrisGPT estimates that at Figure's well-known 2.88s-per-bin pace, only 6,542 packages could be sorted in the same window, making WALL-B 53% more efficient
• The result demonstrates sustained real-world performance of embodied-AI robots in logistics automation
• The comparison draws on each side's separately published data, not a head-to-head test, so real gaps depend on task and environment

2026-08-21 ~ 2026-08-22 · 5 related posts

• WALL-B robot takes on five-hour continuous logistics sorting challenge at WRC 2026 — Scobleizer · 2026-08-21
• [source] WALL-B robot sorts packages faster than Figure 03 — Scobleizer · 2026-08-21
• X Square's WALL-B robot sorts 10,000 parcels in 5 hours — Scobleizer · 2026-08-21
• [source] X Square's WALL-B Outperforms Figure by 53% in Parcel Sorting — ChrisGPT · 2026-08-21

1 near-duplicate retellings: rohanpaul_ai