The line between science fiction and industrial reality has been blurring for years, but rarely has it been drawn as sharply as in recent financial-sector forecasts. For decades, the humanoid robot was a staple of speculative cinema and literature—a machine built in our own image, designed to walk our streets, climb our stairs, and share our spaces. The question was never whether such machines could exist; it was whether they could ever become practical, affordable, and reliable enough to move beyond the laboratory and into everyday life.
According to analysis from Morgan Stanley, that threshold may be closer than many European industrial observers have assumed. The investment bank's projections suggest that by 2035, as many as 13 million humanoid robots could be working alongside human colleagues across the globe. The same forecast extends further, suggesting that by 2050, the number could rise to one billion. These are not incremental improvements to existing automation; they represent a fundamental shift in how work is organised, where it happens, and who—or what—performs it.
The cost trajectory is equally striking. Morgan Stanley's modelling indicates that the annual cost of owning and operating a humanoid robot could fall to approximately $10,000 by the mid-2030s. That figure places the technology in a familiar economic bracket: roughly comparable to the ongoing cost of owning a car. For businesses that have historically viewed advanced robotics as a capital-intensive investment reserved for high-volume manufacturing lines, this pricing level would open entirely new categories of deployment.
Yet the path to this future is not without obstacles. The source material identifies persistent technological challenges, particularly around energy storage and materials science. Batteries must become more efficient to support longer operating hours, and the durable materials required for increasingly sophisticated machines must be developed and refined. These are not trivial engineering problems; they are the constraints that will determine whether humanoid robots remain a niche product or achieve the mass-market penetration that the financial forecasts anticipate.
The source material also makes clear that humanoid robots are not a distant prospect awaiting invention. They are already here. Machines capable of walking, talking, and manipulating objects are operating in academic institutions, industrial facilities, and—through pilot programmes—in domestic and caregiving environments. The technology has moved from the realm of concept demonstrations to working deployments, and the pace of development shows no sign of slowing.
Key findings
The source material provides several concrete data points and product details that merit close attention from industry observers.
**Market scale and cost projections.** Morgan Stanley's forecast of 13 million humanoid robots by 2035 is the headline figure. The same report projects one billion by 2050. The annual cost of ownership, estimated at $10,000, is positioned as comparable to car ownership—a framing that suggests the technology is expected to reach consumer-level affordability rather than remaining an enterprise-only investment. It is important to note that the source does not disclose the methodology behind these projections, nor does it specify whether the figures refer to units sold, units in active service, or cumulative production. What is stated is the scale of the anticipated transformation.
**Current commercial offerings.** The source highlights two specific humanoid robot platforms that are already available or in advanced development.
The first is the GR-2, produced by Fourier. The source describes it as "the world's first mass-produced humanoid robot," though it does not specify what production volume qualifies as mass production. The GR-2's hands are reported to be twice as dexterous as those on its predecessor, the GR-1. The robot can walk at speeds up to 5 mph and is capable of adjusting its grip in real time to suit different tasks and objects—a capability the source characterises as tactile handling. Current deployments are concentrated in academic and industrial settings, with trials underway to evaluate the platform for domestic and caregiving applications. The source lists a starting price of $16,000, though it does not disclose what configuration or feature set that price includes, nor whether it covers software, maintenance, or support.
The second platform is the H1, a Chinese-designed humanoid that achieved a Guinness World Record in 2024 for the fastest running speed by an electric (non-hydraulic) bipedal humanoid robot, at 7.38 mph. The H1 uses LiDAR sensors to maintain a 360-degree perception of its environment and is capable of walking, running, and performing backflips using electric actuators alone. Notably, the source states that the H1 is built on open-source technology, a design choice intended to make the platform accessible to businesses with limited budgets. The source does not disclose the H1's price, nor does it specify which components are open-source and which remain proprietary.
**Technological constraints.** The source explicitly identifies two ongoing challenges: battery efficiency and durable materials. These constraints are presented as necessary areas of innovation to support "increasingly advanced models and use cases." The source does not provide specific performance metrics for current battery life, nor does it disclose the expected operational lifespan of the robots under real-world conditions. What is clear is that the industry's growth trajectory depends on solving these fundamental engineering problems.
**Deployment environments.** The source indicates that humanoid robots are already operating in academic, industrial, domestic, and caregiving settings. The academic and industrial deployments are described as established, while domestic and caregiving applications are at the trial stage. This distinction matters for European operators considering adoption: the technology has proven itself in controlled environments, but the transition to unstructured, human-centric settings is still being evaluated.
What it means for European operators
For European businesses and public institutions, the Morgan Stanley projections and the current state of humanoid robot development raise strategic questions that warrant careful consideration. The source material does not provide region-specific data, so the following analysis is framed around what is known and what remains undisclosed.
**Cost structure and total cost of ownership.** The $10,000 annual cost figure is the most consequential data point for European operators. If this projection holds, it would place humanoid robots within the capital expenditure range of a mid-sized industrial vehicle or a specialised piece of lab equipment. However, the source does not disclose what the $10,000 includes. It is not stated whether this figure covers the amortised purchase price, maintenance, energy consumption, software licensing, insurance, or a combination of these factors. European operators should treat this figure as a directional indicator rather than a budgeting baseline. The $16,000 starting price for the GR-2 suggests that hardware costs are already approaching accessible levels, but the total cost of ownership—including the batteries, materials, and ongoing maintenance that the source identifies as challenges—remains undisclosed.
**Workforce integration.** The source describes a future in which humanoid robots "work alongside humans," not replace them. This framing has significant implications for European labour markets, where automation debates are often more prominent than in other regions. The source does not provide data on job displacement, new job creation, or the skills required to manage humanoid robot fleets. European operators considering adoption will need to assess their own workforce dynamics, including training requirements, safety protocols, and the potential for human-robot collaboration in roles that have historically been resistant to automation.
**Sector-specific opportunities.** The source's description of current deployments provides a useful map of where humanoid robots are gaining traction. Academic and industrial settings are the established markets, which suggests that European research institutions and manufacturing facilities are likely early adopters. The trial deployments in domestic and caregiving settings are particularly relevant for European operators, given the region's demographic trends and the growing demand for elder care and assisted living services. However, the source does not disclose the results of these trials, nor does it specify which caregiving tasks the robots are being evaluated for. European operators in the healthcare and social services sectors should monitor these developments closely, but the evidence base for caregiving deployments is not yet public.
**Technology selection and vendor risk.** The two platforms highlighted in the source represent different approaches to humanoid robot design. The GR-2 is positioned as a mass-produced, commercially available product with a disclosed price point. The H1 is an open-source platform that prioritises accessibility and has demonstrated athletic capabilities—running, backflips—that suggest a focus on mobility and environmental perception. European operators will need to evaluate these platforms against their specific use cases. The source does not provide comparative performance data, reliability statistics, or service and support information for either platform. Vendor lock-in, spare parts availability, and software update policies are not addressed. These are material considerations for any procurement decision, and their absence from the source material should be noted.
**Infrastructure and energy requirements.** The source identifies battery efficiency as a key technological challenge. For European operators, this raises questions about charging infrastructure, energy costs, and operational uptime. A humanoid robot that requires frequent recharging may not be suitable for continuous shift work, and the source does not disclose battery life, charging times, or the energy consumption of either platform. European facilities considering humanoid robots will need to plan for power availability, charging stations, and the potential impact on their energy budgets.
**Regulatory and standards landscape.** The source does not address regulatory considerations, but European operators will be acutely aware that the region's regulatory environment for robotics and AI is evolving. The source does not disclose any certifications, safety standards, or compliance requirements for the GR-2 or H1. Until such information is available, European operators should assume that humanoid robot deployments will be subject to existing workplace safety regulations, data protection laws, and product liability frameworks. The absence of disclosed standards does not mean they do not exist; it means they are not addressed in the source material.
**Timeline realism.** The Morgan Stanley projections extend to 2035 and 2050, which are long-range forecasts subject to significant uncertainty. The source does not disclose the assumptions underlying these projections, nor does it provide a year-by-year adoption curve. European operators should treat the 2035 figure of 13 million units as a plausible scenario rather than a certainty. The current state of the technology—two platforms with limited deployment environments and acknowledged engineering challenges—suggests that the industry is still in its early stages. The gap between today's pilot programmes and a mass-market future is substantial, and the source does not provide evidence that this gap will be closed within the projected timeframe.
**Strategic recommendations.** Based solely on the source material, European operators should consider the following actions. First, monitor the trial deployments in domestic and caregiving settings, as these will likely produce the first evidence of humanoid robot performance in unstructured environments. Second, evaluate the total cost of ownership for the GR-2 and H1 platforms against specific use cases, recognising that the disclosed price points do not include the full range of operational costs. Third, track developments in battery technology and materials science, as the source identifies these as the critical constraints on the industry's growth. Fourth, engage with the open-source ecosystem around platforms like the H1, which may offer European businesses a lower-cost entry point into humanoid robotics. Finally, prepare for a future in which humanoid robots are a routine presence in industrial and service environments, but do so with the understanding that the current evidence base is limited and the projections are directional rather than definitive.
The source material paints a picture of an industry on the cusp of significant expansion, but it also makes clear that the technology is still maturing. The robots that will walk among us by 2035 are not yet fully formed; they are being built today, one engineering challenge at a time.
Sources
https://www.forbes.com/sites/bernardmarr/2025/07/02/13-million-humanoid-robots-will-walk-among-us-by-2035/
Published by Vigla Media OÜ (Estonia).