German robotics manufacturer Neura this week showcased video of humanoid robot, called 4NE-1, performing a variety of activities, like ironing and moving boxes. The promotional video arrives as Nvidia is showcasing a slew of new tools for humanoids at the Siggraph conference in Colorado.
Neura is one of a baker’s dozen of firms that were given early access to Nvidia’s humanoid development and deployment tools . The list also includes 1X, Boston Dynamics, ByteDance Research, Field AI, Figure, Fourier, Galbot, LimX Dynamics, Mentee, RobotEra and Skild AI.
Some, including 1X, Figure and Boston Dynamics, have been among the most prominent names in the emerging form factor. Neura has been teasing the cheekily named 4NE-1 for some time now, but the project has appeared to be in early stages, as the company focused on more traditional form factors in the industrial setting.
When I spoke with Neura CEO David Reger during a humanoid robotics panel at Automate that also included Boston Dynamics, Agility and Apptronik, he promised a closer look at the bot in July. Several of the robots were present on the floor of that event, but none actually functioned, making them little more than photo opportunities for attendees.
This week’s video shows 4NE-1 doing a lot; in some sense, it’s doing more than we’ve seen from other humanoids in the space. Every humanoid needs to be taken with a grain of salt, however — particularly one with a number of short shots and commercial edits. As we’ve noted in the past , these videos in particular don’t represent the system’s capabilities.
Neura, for its part, is using the video to announce its presence on the scene and work with Nvidia’s robotics portfolio. “By combining Neura’s innovative cognitive robotics solutions with Nvidia’s advanced computing power and simulation platforms, we will push the boundaries of humanoid robotics even faster,” Reger said in a comment tied to the news.
The video showcases simulation used to test and train these systems, intercut with shots of the bot in real-world scenarios. Until we’re able to see systems perform repeatable tasks in real-world scenarios at scale, these sorts of videos should always be approached with a degree of skepticism.
As with any company working on a humanoid, I would love to see a lot more raw video of 4NE-1 in the field, doing its thing. Even those humanoids that are out in the field are currently in the pilot stage, with the exception of Agility’s partnership with logistics giant GXO. Even that deal, however, is a ways off from true wide-scale deployment of the kind of magnitude we’ve seen with autonomous mobile robots.
Even so, Nvidia’s commitment to humanoids is a strong vote of confidence for the nascent category. It will also, no doubt, go a long way toward both accelerating development and opening things up to even more new entrants.
“The next wave of AI is robotics, and one of the most exciting developments is humanoid robots,” Nvidia CEO Jensen Huang said in a release. “We’re advancing the entire Nvidia robotics stack, opening access for worldwide humanoid developers and companies to use the platforms, acceleration libraries and AI models best suited for their needs.”