Unitree casts the new H2-X as a robot superhero: confronting a knife-wielding intruder, lifting a car to free a dummy and pushing a child-sized doll out of traffic. These are, of course, staged scenes, not proof of autonomous rescue skills. But quite an action-movie pitch for a humanoid.

Figure 02 might be getting a Terminator-style sendoff. Brett Adcock said Figure was running out of room, and auctioning off the old robots was out: they needed to be destroyed to protect the company’s IP. He ran a poll. “BLOW THEM UP” won with 46.3%. Then Arnold Schwarzenegger replied: “You should melt them.” Adcock’s response? “Arnold! DM sent.” Now he’s posted an AI-generated “Goodbye F.02” teaser with distinctly Terminator 2 imagery. Apparently, an internet poll is no match for a suggestion from the Terminator.

We asked Ma about Delta’s commercial ambitions, why data collection was part of the plan from day one, and what comes next. His answers, plus a closer look at Δ₀’s demonstrations and reported success rates: 👇

Delta Intelligence’s new demo shows a humanoid handling records, making a bed and opening a dishwasher. But homes aren’t its first commercial target. In written answers to Humanoids Daily, founder Xiaojian Ma says the near-term focus is energy and infrastructure—underground facilities, offshore sites and other harsh environments he calls “Mars on Earth.” Delta wants to become an intelligence layer across robot brands. Its approach includes a camera-equipped head and a backpack containing compute and a battery, alongside work with partners on complete humanoids. Ma also explained why Delta built its own D1 data-capture system: learning hand movements alone isn’t enough. Robots need examples of how people balance, kneel, bend and use their whole bodies. The demo shows research progress. Reliable industrial deployment is the next test.

Figure 02 might be getting a Terminator-style sendoff. Brett Adcock said Figure was running out of room, and auctioning off the old robots was out: they needed to be destroyed to protect the company’s IP. He ran a poll. “BLOW THEM UP” won with 46.3%. Then Arnold Schwarzenegger replied: “You should melt them.” Adcock’s response? “Arnold! DM sent.” Now he’s posted an AI-generated “Goodbye F.02” teaser with distinctly Terminator 2 imagery. Apparently, an internet poll is no match for a suggestion from the Terminator.

“Keeping them working has to be as scalable as building them.” We asked O-ID CEO Stian Jakobsen how modular robot maintenance would actually work. The plan: detect wear in an elbow, ship a replacement, then have a trained factory worker power down the robot and swap the module. It would load its stored calibration and run readiness checks before returning to work. The old module goes back for refurbishment. No engineer flying in for every replacement. Why build in Japan? Jakobsen points to precision automotive manufacturers that could both supply the robots and use them—and relationships built by being there. This is the intended workflow, still under development. Read the updated article:

“Keeping them working has to be as scalable as building them.” We asked O-ID CEO Stian Jakobsen how modular robot maintenance would actually work. The plan: detect wear in an elbow, ship a replacement, then have a trained factory worker power down the robot and swap the module. It would load its stored calibration and run readiness checks before returning to work. The old module goes back for refurbishment. No engineer flying in for every replacement. Why build in Japan? Jakobsen points to precision automotive manufacturers that could both supply the robots and use them—and relationships built by being there. This is the intended workflow, still under development. Read the updated article:

Mk1.5 adds audio, object tracking, web search and sub-agents to embodied AI. Up to 4.7× faster than Mk1 in Perceptron's tests. It watches the camera feed and picks which commands a drone, robot or pair of smart glasses should run next.

Feather is launching a ~$30K wheeled humanoid that lets developers choose their own AI. We spoke with CEO Hoa Mai about the company’s bet on simple, reliable hardware—and why robotics may need an “iPod moment” first. Our interview with @featherrobotics 👇

Feather is launching a ~$30K wheeled humanoid that lets developers choose their own AI. We spoke with CEO Hoa Mai about the company’s bet on simple, reliable hardware—and why robotics may need an “iPod moment” first. Our interview with @featherrobotics 👇

“Where were you when the robots came?” Morgan Stanley’s Adam Jonas thinks combining AI and robotics could multiply global GDP eight to ten times “before you and I retire.”

What happens when a factory robot breaks? Tokyo startup O-ID is building modular humanoids with replaceable joints, limbs and compute units. Its pitch: on-site repairs in minutes, supported by sensors that flag component wear before failure. It has raised $1.2 million in pre-seed funding led by TAWANI Ventures, targeting Japanese automotive manufacturing and logistics first. The modularity is intriguing. Measured repair times, reliability results and a customer rollout schedule are still to come.

Tesla’s Android app has contained Optimus renders labeled “gen3” since August 27. We compared four app builds: higher-resolution versions arrived September 5, and the latest release contains the same files. Compared with the gold variant, the “gen3” render covers the exposed mechanisms around the hips and lower legs. A glimpse of the next Optimus? Possibly. These are app assets—not confirmation of the final hardware or a working robot interface.

America’s humanoid robot ambitions face a hardware bottleneck. Apptronik CEO Jeff Cardenas says actuators—the motors, gears and electronics that act as a robot’s muscles—can account for up to 60% of its parts cost. And the U.S. supply gap goes beyond rare earths: “We have major shortages in things like gears where we don’t have the supply base that we need.” Apptronik wants to build that capacity in Texas. Scaling humanoids will require scaling the companies that make their parts, too. 🎥 Fox Business

Skild AI says S1 plays soccer against humans and robots after 140+ years of simulated self-play. “This method scales, and we will scale it,” says CEO Deepak Pathak. How it works—and where human guidance comes in:

Tesla’s Android app includes two versions of this Optimus “gen3” render: a brighter full-body image and a darker one named “inactive.” Could they represent different app states—an inactive robot, or an account without an Optimus linked?Possibly, but that’s speculation. This GIF simply alternates between the extracted images. It is NOT a recording of the app, and the assets don’t confirm a working Optimus interface or the final Gen 3 design.

Tesla’s Android app has contained Optimus renders labeled “gen3” since August 27. We compared four app builds: higher-resolution versions arrived September 5, and the latest release contains the same files. Compared with the gold variant, the “gen3” render covers the exposed mechanisms around the hips and lower legs. A glimpse of the next Optimus? Possibly. These are app assets—not confirmation of the final hardware or a working robot interface.

America’s humanoid robot ambitions face a hardware bottleneck. Apptronik CEO Jeff Cardenas says actuators—the motors, gears and electronics that act as a robot’s muscles—can account for up to 60% of its parts cost. And the U.S. supply gap goes beyond rare earths: “We have major shortages in things like gears where we don’t have the supply base that we need.” Apptronik wants to build that capacity in Texas. Scaling humanoids will require scaling the companies that make their parts, too. 🎥 Fox Business

1X wants to ship 50,000 robots next year. CEO Bernt Børnich says the real challenge is “ensuring that you don’t get them back.” In this interview with @ti_morse, he explains why scaling production means uncovering problems that never appeared in smaller batches—and fixing them before ramping further. The target spans homes, enterprise applications and developers. It isn’t a promise of 50,000 fully autonomous household helpers.