In June 2026, German humanoid robotics company Neura Robotics announced the completion of its Series C funding round, raising up to $1.4 billion and reaching a valuation of approximately $7 billion. This round attracted heavyweight investors including NVIDIA, Amazon, Qualcomm, Bosch, the European Investment Bank, and stablecoin issuer Tether. This is not an isolated funding event. According to Dealroom, since 2026, robotics companies have collectively raised $55.8 billion, setting a new all-time high—nearly double the previous annual record. Capital is pouring into the robotics + AI convergence track at an unprecedented pace and scale.
The industry’s strategic focus on this sector is becoming increasingly clear. NVIDIA CEO Jensen Huang has summarized the evolution of AI technology into three generations: from Perception AI to Generative AI, then to Agentic AI, with the next frontier being Physical AI—AI that can operate, reason, plan, and act in the real world. Amazon Web Services, MassRobotics, and NVIDIA have jointly launched the Physical AI Fellowship accelerator program, with the second cohort in 2026 now open to robotics startups worldwide. Judging by both the scale of capital inflows and the strategic moves of leading tech companies, Physical AI has moved beyond proof-of-concept and is on the cusp of large-scale deployment.
The Scope, Scale, and Market Structure of Physical AI
At its core, Physical AI aims to enable AI to break free from the digital realm and operate within the real, physical world. According to MarketsandMarkets, Physical AI refers to the integration of artificial intelligence into robots, autonomous vehicles, drones, industrial equipment, and other physical systems, empowering these systems to perceive, analyze, and interact with the real world. Unlike traditional AI, which generates text or images, Physical AI’s output is tangible—objects are moved, assembled, or transported, resulting in real-world physical actions. A deep-dive report from Zheshang Securities notes that Physical AI must answer two fundamental questions: How will the world change next, and how will the world respond after a physical entity acts? This technical capability underpins the three core scenarios of autonomous driving, embodied intelligence, and industrial software.
Market size estimates vary widely depending on methodology, but there is strong consensus on the growth trajectory. MarketsandMarkets, using a focused statistical approach, projects the global Physical AI market to grow from $1.5 billion in 2026 to $15.24 billion by 2032, with a compound annual growth rate of 47.2%. Using a broader definition—including all AI-enabled physical systems such as industrial robots, autonomous vehicles, surgical robots, military automation, and smart infrastructure—the global market is estimated at around $383 billion in 2026, potentially reaching $3.26 trillion by 2040. A more macro perspective from hedge fund Coatue Management estimates the total Physical AI market at a minimum of $6 trillion, about 50% larger than the pure digital AI market. At CES 2026, Jensen Huang went even further, stating that Physical AI could reshape the $50 trillion manufacturing and logistics industries. Despite significant differences in these estimates, they all point to the same conclusion: the market size of Physical AI is moving from the tens of billions toward the multi-trillion-dollar range.
Demand-side pressures are equally significant. About 2.5 billion people worldwide are engaged in various forms of physical labor, generating roughly $50 trillion in annual GDP. With accelerating population aging, labor shortages in manufacturing, logistics, and healthcare continue to widen. Meanwhile, the costs of sensors, cameras, and robotics-grade processors are rapidly declining, and the technical maturity of Generative AI and Agentic AI is rising. Together, these factors are creating the structural momentum for Physical AI’s adoption. As both supply and demand conditions align, large-scale capital inflows are a natural result for the industry.
Competitive Landscape and Product Differentiation Among Global Physical AI Companies
Neura Robotics’ $1.4 billion funding round is noteworthy not just for its size, but for what it reveals about the competitive landscape: the Physical AI sector now features multiple tiers and parallel technology paths. Public data shows the world’s top-funded humanoid robotics companies include: Figure AI, with total funding of about $1.75 billion and a recent valuation of $39 billion; UBTECH, with $940 million raised; Apptronik, with about $1 billion raised and a $5 billion valuation; Agility Robotics, with $330 million raised and a valuation between $1 billion and $1.75 billion; and Neura Robotics, now valued at $7 billion after this round. Additionally, Boston Dynamics continues to advance the commercialization of its Atlas electric humanoid within the Hyundai Motor Group.
These companies differ significantly in their technical approaches and business models. Figure AI, founded by serial entrepreneur Brett Adcock in 2022, has rapidly expanded using a VC-heavy model, attracting investments from NVIDIA, Microsoft, the OpenAI Startup Fund, and Amazon founder Jeff Bezos in its Series B round. Its Figure 03 home robot is priced at around $20,000 and targets the consumer market. Apptronik operates via an industry alliance model, raising about $1 billion and forging strategic partnerships with Google DeepMind, GXO Logistics, and Mercedes-Benz. Its Apollo robot is designed as a general-purpose platform, featuring both bipedal and wheeled configurations, and is preparing for mass production in Texas and California. Agility Robotics focuses on logistics, with its Digit humanoid robot already in pilot deployment within Amazon’s warehouse system, and has attracted investment from Amazon, NVIDIA, and SoftBank. Boston Dynamics represents another path—after Hyundai acquired an 80% stake for $880 million, it is leveraging the automaker’s resources to drive commercialization.
China’s market has also developed a clear, multi-tiered competitive structure. Currently, there are over 200 humanoid robotics concept stocks listed on the A-share market, with a combined market cap exceeding 6.1 trillion yuan. Unitree Robotics’ STAR Market IPO has passed review and entered the registration phase, making it likely to become the first humanoid robotics stock on the A-share market in Q3. On the Tesla supply chain front, Optimus V3 is set to begin large-scale production in summer 2026, while BYD has officially announced its entry into the humanoid robotics sector with project "Yao Shun Yu," planning to deploy 20,000 units in its own factories in 2026. The first phase of the Xi’an Robotics Industrial Park is operational, with an annual capacity of 50,000 units and a target price below 200,000 yuan per unit. From a supply chain perspective, upstream and downstream players such as Midea Group, Shenghong Technology, Lens Technology, Inovance Technology, and Ganfeng Lithium have all deeply engaged in the humanoid robotics track.
NVIDIA’s pivotal role in the Physical AI ecosystem is especially noteworthy. As the global leader in GPUs and edge computing chips, NVIDIA’s Isaac GR00T development platform has become an industry-standard foundation. NVIDIA and Unitree Robotics have jointly launched the first humanoid robot reference design, H2 Plus, built on the open-source Isaac GR00T platform. NVIDIA also announced its next-generation Feynman chip, specifically designed for Physical AI and expected to launch in 2028. This three-tier structure—chip, algorithm, and platform—positions NVIDIA as an infrastructure provider for Physical AI, mirroring Amazon Web Services’ strategy of supporting the Physical AI startup ecosystem through AWS compute resources. In March 2026, Neura Robotics and AWS announced a strategic partnership to globally scale the Neuraverse platform via AWS.
Neura Robotics as a Case Study: A Sample Analysis of Physical AI Investment Logic
A $1.4 billion funding round, a $7 billion valuation, and participation from over 10 top-tier institutions—Neura Robotics’ Series C is one of the most iconic single deals in Physical AI for 2026. Analyzing this case helps clarify the core logic driving industrial capital’s investment decisions in Physical AI.
From a technology perspective, Neura Robotics employs a multi-form product strategy. Its portfolio includes the 4NE1 humanoid robot, the consumer-grade wheeled robot MiPA, and the MAV warehouse transport series, all powered by the AURA AI navigation system. The advantage of this multi-product strategy is that it allows the company to deploy various physical systems on a unified AI platform, accumulating real-world operational data across industrial, logistics, and consumer markets. This creates a data feedback loop that accelerates algorithm iteration. According to its official website, the company will use this funding for three main purposes: global deployment of humanoid robots, expansion of production and delivery capabilities, and development of next-generation Physical AI systems. These objectives align with the three critical thresholds Physical AI companies must cross: from "technology validation" to "scaling up" and ultimately to "paradigm upgrade."
From a capital structure standpoint, Neura Robotics’ investor base is highly diversified, including Qualcomm at the chip layer, Amazon and NVIDIA at the technology layer, Bosch and Schaeffler in industrial infrastructure, and even Tether, which operates under a unique regulatory regime. Notably, Tether’s investment is purely equity-based, with no blockchain protocol or token issuance attached. This demonstrates that institutional investors’ interest in Physical AI has moved beyond hype and into substantive industrial capital allocation. The intensity of this cross-sector capital collaboration further confirms that Physical AI is evolving from a single hard-tech track into a multi-industry integration platform.
However, this funding round also presents significant risks that must be recognized. First, whether the full $1.4 billion will be received depends on Neura Robotics achieving certain predefined milestones. This means the company must deliver on hard metrics such as mass production capacity, order fulfillment, and commercialization progress to secure the full amount. Second, Neura previously raised about $55 million in 2023, and its funding has soared from $55 million to over $1 billion in just three years. While this reflects rapid sector growth, it also increases market expectations for valuation and product delivery. Third, the humanoid robotics field still faces intense homogeneity risk—many leading companies have overlapping technologies, application scenarios, and target customers. Building and validating differentiation will require more operational data.
Investment and Risk Assessment Framework for the Physical AI Sector
Based on the above, Physical AI as an investment sector can be summarized into three interconnected logic layers.
The first layer is infrastructure. Chips and computing power are the core foundation for Physical AI. NVIDIA, with its first-mover advantage in GPUs and ecosystem barriers in robotics software platforms, dominates this layer. Chipmakers like Qualcomm are entering through edge computing SoCs. The hardware segment of the Physical AI market accounted for the largest share in 2025–2026. Investment logic here is relatively mature, but competition is stabilizing, with incremental growth mainly driven by expanding downstream application scenarios and the resulting spillover in compute demand.
The second layer is the robotics body and platform. This is currently the most heavily funded area globally, including Figure AI, Apptronik, Agility Robotics, Boston Dynamics, UBTECH, and Unitree Robotics. This layer combines hardware manufacturing with software algorithms, setting the highest investment threshold and greatest divergence in technical approaches. Differentiation centers on three aspects: mechanical design (bipedal vs. wheeled vs. hybrid), AI decision system architecture (centralized vs. distributed), and scenario entry points (industrial logistics vs. home service vs. public safety). There is no clear evidence yet that any single technical route holds a decisive lead, so investing at this layer requires evaluating both engineering and algorithmic capabilities—a lead in just one area is not enough to create a lasting moat.
The third layer is industry solutions and data services. This involves providing end-to-end automation solutions for specific scenarios based on the underlying chips and platforms, as well as leveraging real-world data collected during robot operations. The Physical AI Fellowship accelerator, co-launched by AWS and NVIDIA, is an early initiative in this layer—by providing technical and computing resources, it helps Physical AI startups worldwide overcome early-stage R&D barriers. The investment logic here is closer to the SaaS model, but the commercial maturity of these solutions will take more time to prove.
Alongside investment opportunities, the Physical AI sector faces several risks that require careful assessment. The first is the high uncertainty of mass production for humanoid robots. According to GF Securities, 2026 is a key inflection point for moving from "zero to one" in humanoid robotics, but scaling from "thousands" to "tens of thousands" of units has yet to be validated in terms of supply chain stability, quality control, and cost curves. Second, there is the risk of convergence in technical approaches. The two main routes—world models and vision-language-action models—are both in the early, unconverged stages, and a breakthrough in either could render large prior investments obsolete. Third, there are security boundary concerns for Physical AI systems. Deepu Talla, NVIDIA’s VP of Robotics, emphasizes that Physical AI development and deployment spans the entire lifecycle from data generation, training, and simulation to secure deployment. Any failure at any stage could lead to irreversible physical consequences—making the rollout of Physical AI inherently slower than pure digital AI. Additionally, global macroeconomic volatility, geopolitical impacts on supply chains, and regulatory changes in major economies all represent key variables affecting the valuation framework for the Physical AI sector.
Conclusion
From the $1.4 billion joint investment by NVIDIA and Amazon to the global annual robotics fundraising total of $55.8 billion, the Physical AI sector now stands at the tipping point of explosive growth, driven by both capital and industry. What makes this sector unique is that it not only embodies the paradigm shift of AI from the digital to the physical world, but also involves the deep integration of mature industries such as semiconductor chips, sensor manufacturing, motion control systems, and industrial automation.
For investors and researchers, understanding the essence of Physical AI—enabling AI to perceive the world, reason about causality, and execute physical actions in a closed loop—is the foundation for building an effective analytical framework. On this basis, tracking the technical differentiation, commercialization progress, and capital structure changes of leading companies is key to identifying investment timing and industry inflection points. Whether Physical AI’s ultimate commercial scale will reach the multi-trillion-dollar levels projected by Coatue and Jensen Huang depends on breakthroughs in technical maturity, mass production feasibility, and safety. But at this moment, one thing is clear: Physical AI is no longer a distant sci-fi fantasy—it is on the verge of large-scale deployment and becoming an industrial reality.
