The Robotics Companies No One's Talking About (Yet)

Humanoid robotics is gradually entering the mainstream conversation. Every few weeks there is another viral demonstration, another funding announcement, or another prediction about robots entering the workforce. The conversation has become louder, but it is still focused on a small group of companies.

In reality, humanoid robotics has developed into a far broader ecosystem. Across Europe, North America and Asia, companies are tackling entirely different pieces of the puzzle. Some are building complete humanoid platforms. Others are developing the intelligence that allows robots to understand the physical world. Some are rethinking actuation, dexterity and mobility, while others are solving the manufacturing challenges that determine whether these machines can ever be produced economically.

The pace of development has accelerated considerably over the past eighteen months, and the discussion is now geared towards understanding where humanoids create the greatest economic value.

Many of the companies driving that progress receive remarkably little attention.

The Names to Know

Figure AI, Tesla Optimus, Unitree and 1X are among the companies that have brought humanoid robotics into the mainstream. They have become the reference point for how most people understand humanoids today.

Just below them is another group of companies making significant progress. They may not generate the same level of attention, but they are steadily building commercial traction, securing strategic partnerships and developing technologies that could prove just as influential over the coming decade. Apptronik and NEURA Robotics belong firmly in this category.

Apptronik

Apptronik is one of the industry’s strongest commercial contenders.

Founded out of the Human Centered Robotics Lab at the University of Texas, Apptronik spent more than a decade developing robotic systems before introducing Apollo, its general purpose humanoid. The company’s early work included robotics projects for NASA, giving the team experience building machines designed to operate alongside people in demanding environments.

Apollo is being developed for manufacturing and logistics, where labour shortages and repetitive tasks make humanoids commercially attractive. Mercedes-Benz has been evaluating Apollo within its production facilities, while Google DeepMind is collaborating with Apptronik to advance embodied AI for humanoid robots. Earlier this year, the company introduced Apollo 2 alongside Robot Park, a dedicated training facility designed to collect large scale operational data and accelerate robot learning.

With established industrial partnerships, an experienced engineering team and a growing focus on manufacturing, Apptronik has positioned itself among the leading commercial humanoid companies.

NEURA Robotics

While many robotics startups remain focused on developing a single product, Germany’s NEURA Robotics has been building an entire cognitive robotics ecosystem.

The company first established itself through collaborative robots before expanding into mobile robotics and, more recently, humanoids. Rather than treating these as separate product categories, NEURA believes every robot should learn through a shared cognitive platform, allowing improvements made by one machine to benefit every other system connected to its software ecosystem.  

In recent weeks, the company secured a funding round of up to $1.4 billion backed by organisations including Nvidia, Amazon, Qualcomm, Bosch and the European Investment Bank. NEURA says it now has an order backlog exceeding $1 billion as it prepares to scale production across industrial, logistics and service applications.  

Europe has often been viewed as lagging behind the United States and China in humanoid robotics. NEURA is one of the few companies demonstrating that it can compete on both technology and scale.

Companies To Watch

Clone Robotics

Few companies are questioning the physical design of humanoids as fundamentally as Clone Robotics.

Most humanoid developers rely on electric motors, harmonic drives and conventional actuators. Clone Robotics designs  artificial muscles and tendon driven systems that more closely resemble human anatomy. The objective is not simply to build robots that look human, but to reproduce the efficiency, flexibility and dexterity of biological movement itself.

It is a high risk engineering challenge, but one with significant long term implications. Dexterity remains one of the biggest barriers to general purpose humanoids, particularly when manipulating unfamiliar objects. If Clone’s biomimetic approach proves commercially viable, it could influence how future humanoids are engineered rather than simply how they are programmed.

UBTech

Outside China, UBTech rarely receives the attention it deserves.

Founded in 2012, the company is one of the world’s longest established commercial robotics businesses and has spent more than a decade building manufacturing capability, supply chains and enterprise relationships. While many newer startups are still preparing for large scale production, UBTech is already deploying humanoid robots into industrial settings across China.

The company already has years of manufacturing experience, established enterprise customers and an existing robotics business. As China continues investing heavily in humanoids, those foundations could allow it to scale more quickly than many newer entrants.

FieldAI

Building a humanoid robot is only half the problem. Teaching it to operate safely in an unpredictable world is arguably the more difficult challenge.

FieldAI is developing foundation models that allow autonomous machines to perceive unfamiliar environments, make decisions with incomplete information and adapt to situations they have never previously encountered. Rather than programming robots task by task, the company is attempting to build a more general intelligence layer capable of transferring across different robotic platforms.

Although FieldAI’s software extends well outside humanoids into construction, defence and industrial robotics, the underlying technology addresses one of the major challenges facing the entire sector. General purpose robots will only become commercially viable once they can operate reliably outside carefully controlled environments.

Physical Intelligence

One of the biggest unanswered questions in robotics is whether a single AI model can learn skills that transfer naturally across different robots, environments and tasks.

Physical Intelligence believes that is possible.

Its π foundation models are designed to learn from diverse forms of interaction rather than being programmed for individual behaviours. The long term objective is to create an intelligence layer capable of controlling many different robotic systems in much the same way large language models now power many different software applications.

If this approach works, the value of a robotics company may not come only from the robot it builds, but from the intelligence layer that can operate across many different machines. Physical Intelligence is one of the companies trying to build that layer.

Proception AI

Precision remains one of the final barriers separating humanoid robots from human capability.

Picking up a cardboard box is relatively straightforward. Fastening a zip, folding laundry or handling fragile objects requires an entirely different level of perception and control. These tasks depend on touch as much as vision.

Proception AI is developing robotic hands equipped with tactile sensing designed to give machines a far richer understanding of the objects they interact with. By combining touch feedback with increasingly sophisticated manipulation algorithms, the company aims to close one of the largest remaining gaps between human dexterity and robotic precision.

As humanoids move into homes, healthcare and advanced manufacturing, technologies like these may prove just as important as improvements in locomotion or artificial intelligence. The ability to manipulate the physical world with confidence is ultimately what transforms a robot from an impressive machine into a useful one.

Bullish on Robotics? So Are We.

XMAQUINA is a decentralized ecosystem, giving a global community early exposure to the world’s leading robotics companies before they disrupt trillion-dollar industries.

Now, you don’t have to be sidelined. Own the rise of humanoids.

Join our Discord and connect with thousands of futurists building the XMAQUINA DAO.

Follow us on X for the latest updates.