NVIDIA, Cartwheel Robotics And Ground Control Robotics Make Waves In Robotics

Very few technologies have emerged and evolved as quickly as robotics. We have gone from automated vacuum cleaners to humanoid robots in just a few decades.

Now, robotics is entering the era of "Physical AI," where intelligent machines interact autonomously with the real world. This new branch of robotics is driven by AI innovations and advances in computing, a shift that is set to transform industries and daily life.

Companies like NVIDIA, Cartwheel, and Ground Control Robotics are leading the way, equipping developers with the tools to build more capable, adaptable humanoid robots.

So, here’s the latest from the world of next-gen robotics!
 

NVIDIA Advances AI-Powered Humanoid Robotics

NVIDIA is accelerating humanoid robot development with its cloud-to-robot platforms, unveiling major updates, especially Isaac GR00T N1.5 for advanced reasoning and GR00T-Dreams for generating synthetic motion data.

These innovations, coupled with the NVIDIA Blackwell systems, are accelerating the pace of robotic development across industries.


Top robotics developers, including Agility Robotics, Boston Dynamics, Fourier, Foxlink, Mentee Robotics, NEURA Robotics, and XPENG Robotics, are already leveraging NVIDIA’s Isaac platform to advance their humanoid robot projects.

As Jensen Huang, founder and CEO of NVIDIA, stated, "Physical AI and robotics will bring about the next industrial revolution. From AI brains for robots to simulated worlds to practice in or AI supercomputers for training foundation models, NVIDIA provides building blocks for every stage of the robotics development journey.”

A critical challenge in developing highly skilled robots is the immense amount of diverse data required for training. However, NVIDIA is addressing this data gap with its new Isaac GR00T Data Generation Blueprint.

Next, GR00T-Dreams, showcased at COMPUTEX, will enable the generation of vast amounts of synthetic motion data, or "neural trajectories." Developers can use this platform to teach robots new behaviors, including adapting to changing environments, by post-training Cosmos Predict world foundation models (WFMs) for their robots.

The efficacy of this approach is evident in the development of GR00T N1.5 from NVIDIA Research, which utilized the GR00T-Dreams blueprint to develop the model in just 36 hours—what would’ve taken nearly three months.

Further, GR00T N1.5 offers better adaptability, improved object recognition via user instructions, and higher success in material handling. Early adopters like AeiRobot, Foxlink, Lightwheel, and NEURA Robotics are already using it to boost flexibility, validate synthetic data, and fast-track automation.

NVIDIA is also streamlining robot training with tools like Cosmos Reason on Hugging Face, GR00T-Mimic for scalable motion data, and a 24,000-trajectory Physical AI Dataset. Isaac Sim 5.0 and Isaac Lab 2.2 will soon be open-sourced, helping companies like Foxconn and Boston Dynamics in simulation, training, and validation.

To meet the computing demands of robotics, manufacturers can use RTX PRO 6000 systems, while NVIDIA Blackwell and DGX Cloud will enable large-scale training, deployable on Jetson Thor for faster on-robot inference and performance.

With NVIDIA laying the technical foundation, innovators like Cartwheel Robotics are now redefining what humanoids can be, shifting their purpose from factory floors to friendly, everyday companions.
 

Cartwheel Robotics Shifts Focus To Home Companions

While most humanoid robotics today focuses on industrial tasks in warehouses, manufacturing plants, and factories, Cartwheel Robotics is taking a distinctly different path.


Founder Scott LaValley envisions humanoid robots that foster connection and bring "joy, warmth, and a bit of everyday magic" into homes, aiming for general-purpose home companionship rather than mere utility.

With ten years at Boston Dynamics (Atlas) and five at Disney (Baby Groot), LaValley brings a unique perspective on human-robot interaction.

He recounts how his children’s fearful reaction to Boston Dynamics' "terrifying robots" contrasted sharply with their joyful connection to Baby Groot, a pivotal experience that shaped his vision for Cartwheel.

"At Boston Dynamics, we were known for terrifying robots," LaValley remembers. "At Disney, I brought my kids in and they would light up with a big smile on their face and ask, ‘Is that really Baby Groot? Can I give it a hug?’ And I thought, this is the type of experience I want to see robots delivering."

LaValley added, “These humanoid robots are built to be tools. They lack personality. They’re soulless. But we’re designing a robot to be a humanoid that humans will want in their day-to-day lives.”

Cartwheel’s prototype, Yogi, features “toddler-like” proportions and a friendly design to feel more like a character than a machine. A simpler and more customizable version, Speedy, also exists to target broader markets.

Creating a viable social robot has proven to be tough, with past efforts faltering on cost and novelty. Cartwheel has now raised the bar by choosing a complex humanoid form, demanding advanced motion, balance, and child-safe design.

However, LaValley believes that advancements in AI, particularly for whole-body motion generation, can address the challenge of maintaining long-term novelty by making each robot's behavior unique.

Cartwheel aims to bring friendly humanoids into homes, likely via a rental model—though LaValley admits justifying the cost will take time.

"This problem won’t be solved in the next year, or maybe not even in the next five years," he admits. Cartwheel Robotics, having bootstrapped development through robotics engineering services, plans to deliver several dozen Speedy variations to museums and science centers over the next year.

Their "north star" remains the general-purpose home robot, a vision that requires continuous development and real-world partnerships.

While Cartwheel reimagines humanoids for the home, Ground Control Robotics is taking a different path with a new wave of intelligent physical machines built for the fields. Read on!
 

Ground Control Robotics Launches Bio-Inspired Centipede Robots For Farms

Shifting from humanoids to the unconventional, Ground Control Robotics (GCR), founded by Georgia Tech’s Dan Goldman, is using bio-inspired robotic centipedes to transform agricultural crop management.

This novel approach moves beyond the typical snake-like robots seen in disaster relief, finding a unique and commercially viable niche in agricultural environments. GCR’s robots feature a sensor-equipped head and identical motorized segments, offering flexibility with simple control methods.


Goldman explains, "Centipede robots, like snake robots, are basically swimmers." The addition of legs allows these "swimmers" to navigate a wider range of environments, with a coordinated lifting and lowering pattern of the legs generating a fluid-like thrust.

"When you tune things properly, the robot goes from being stiff to unidirectionally compliant. And if you do that, what you find is almost like magic—this thing swims through arbitrarily complex environments with no brain power," he notes.

These robots are built for tough farm terrains, such as steep, rocky slopes with crops like grapes, where traditional machines struggle and farmers still rely on costly manual labor.

GCR estimates weed control costs over $300 per acre for blueberries and over $1,000 for strawberries. With labor shortages and the push to cut pesticides, their robotic centipedes offer a strong alternative.

GCR’s multi-legged design lets their robots move very close to crops without damage, being an important advantage in tight, cluttered environments compared to quadrupeds or wheeled robots.

Goldman emphasizes, "We want to send the robot as close to the crops as possible. And we don’t want a bigger, clunkier machine to destroy those fields." However, GCR's insight is that "it’s possible to generate reliable motion without any sensing at all, if we have a lot of legs." This redundancy makes the robotic system inherently robust and adaptable.

A key benefit is cost-effectiveness, with each robot expected to cost around $1000, along with inexpensive leg modules and mechanical intelligence to reduce the dependency on sensors and real-time computing.

Its vision is for decentralized swarms of robots to scout and remove weeds 24/7. In fact, Ground Control Robotics is piloting with Georgia blueberry and vineyard owners to refine the robot’s mobility and sensing capabilities.

While broader applications like disaster relief or military uses are being explored, GCR's initial focus on agriculture shows a clear path from research to practical, impactful robots.

These advancements, focused on both social connection and real-world tasks, highlight a future where AI-powered robots (humanoid or not) become essential partners in the daily lives at home, workplaces, farms, and industrial settings.

Do you think robots can become everyday helpers in both our homes and offices? How soon do you believe AI humanoids will become a part of our lives?

Let us know your thoughts in the comments below!