Another humanoid crosses 200‑hour continuous uptime, Market Maker reports

Market Maker_HQ said on May 25 that “another humanoid” had crossed 200 hours of continuous uptime, framing the milestone as evidence that the sector is being judged less on short demos and more on whether machines can stay on task for days. The post tied that claim to spatial intelligence, decentralized infrastructure and edge computing, and cited Auki Labs as an example of shared machine perception. The post did not name the robot company in the text visible from secondary reporting, but a separate report published on May 22 said Figure AI had completed a 200-hour autonomous package-sorting run with its Figure 03 system. ### Why does the 200-hour figure matter more than the old 8-hour benchmark? Figure AI’s reported test began as an eight-hour livestream and was later extended to 200 hours, according to Rockingrobots, which cited company statements and comments from Chief Executive Brett Adcock. The setup was presented as a reliability and endurance test in a logistics workflow rather than a one-off stunt. (rockingrobots.com) The difference is operational duration. An eight-hour run can show that a robot can finish a shift. A 200-hour run tests whether a system can keep working through charging, handoffs, resets and repeated object handling over multiple days, according to the same report. ### Was one robot running for 200 hours straight? Rockingrobots reported that the 200-hour result used multiple Figure 03 units rather than one robot operating continuously for the full period. (rockingrobots.com) During the livestreamed setup, robots rotated in and out of the work cell while others charged, the report said. That distinction matters because the uptime claim is about system availability, not a single battery cycle. (rockingrobots.com) In warehouse terms, the test was described as a fleet problem involving charging, recovery and handover, not just whether one machine could stay powered indefinitely. ### What exactly were the robots doing during the run? (rockingrobots.com) Figure AI’s reported task was narrow: identify packages, pick them up and place them on a conveyor with the barcode facing down. Rockingrobots said Figure described the system as autonomous, powered by its Helix-02 AI model, with automatic reset behavior when the robots encountered situations outside their normal operating range. (rockingrobots.com) The same report said the work cell was controlled and repetitive, and it cautioned that the result did not show general-purpose warehouse readiness. That limitation is important because long uptime on a fixed task is different from broad deployment across varied workflows. ### Why did Market Maker_HQ bring up spatial intelligence and Auki Labs? (rockingrobots.com) Auki says its network is intended to make the physical world “browsable, navigable and searchable” for AI, and describes its system as an external sense of space that machines can use to understand the world together. On its website and GitHub, Auki calls the posemesh a decentralized machine-perception network and collaborative spatial-computing protocol. (rockingrobots.com) Market Maker_HQ’s reference to shared perception fits that framing. If robots are expected to operate for days and coordinate across spaces, they need maps, localization and common scene understanding that can persist beyond a single robot’s onboard sensors. That connection is an inference from Auki’s published description and the Market Maker_HQ post’s emphasis on decentralized spatial infrastructure. ### What does this say about where humanoid testing is heading? (auki.com) Figure’s public materials show an increasing focus on production, logistics and full-body autonomy. Its news page lists “Ramping Figure 03 Production” on April 29, 2026, and “Helix 02: Full-Body Autonomy” on January 27, 2026, alongside earlier logistics updates. The next concrete checkpoints are likely to be additional production and autonomy updates from Figure, and further technical releases from Auki around its posemesh tools and documentation, both of which remain publicly available on their official sites. (auki.com) (figure.ai)