The humanoid robotics sector is undergoing a significant narrative shift. For years, the dominant framing in industrial and technology media has positioned these machines as solutions to labor shortages, factory automation, and logistics efficiency. The pitch was straightforward: deploy a humanoid robot where a human-shaped worker is needed, and solve the problem of unfilled shifts and repetitive tasks.
That framing remains valid, but it is no longer the only story. A growing body of evidence, market projections, and product development activity points to a parallel and potentially larger use case: humanoid robots as companions, social partners, and even friends. This is not a niche curiosity or a science-fiction fantasy. It is a market segment with a stated valuation in the hundreds of billions, driven by demographic trends, cultural attitudes, and a measurable global loneliness crisis.
The source material for this analysis, published by Forbes in 2025-01, examines this transition through the lens of Realbotix’s Aria humanoid, the experiences of elderly users in the United States, and the broader market trajectory for humanoid robots through 2035. For European operators—whether they run care facilities, robotics startups, or industrial automation divisions—this dual-use reality demands attention. The same platform that might one day stock a warehouse shelf could also sit in a living room. Understanding the scope, the limits, and the data behind this shift is essential for strategic planning.
This analysis re-examines the source material’s key claims, places them in context, and draws implications for European businesses and public-sector operators. All figures and statements are drawn exclusively from the provided source text. Where the source is silent, this analysis says so explicitly.
Key findings
The source material presents several distinct but interconnected findings. First, the humanoid robot market is still in its infancy in terms of actual deployments, but the projected growth curve is steep. According to the Forbes report cited in the source, global humanoid robot shipments totaled 13,317 units in the most recent full year. That is a small number—barely a rounding error compared to traditional industrial robot installations. However, the same report projects that figure will grow to 2.6 million units by 2035. That implies a near-doubling of shipments annually over the next decade.
The source is careful to flag a critical caveat: the report itself acknowledges uncertainty about how many of those 13,317 units represent commercial sales versus demonstration models or pilot deployments. This is not a minor footnote. It underscores that the industry is still in an early-stage, pre-commercial phase for many players. A unit shipped to a trade show floor is not the same as a unit sold to a paying customer who relies on it daily. European operators should treat the 2.6 million figure as a directional indicator of investor and manufacturer intent, not as a confirmed order book.
Second, the source material highlights a specific product that embodies the companion-robot trend: Realbotix’s Aria. Aria is described as a hyper-realistic humanoid robot designed as a companion, friend, and “maybe more.” The physical platform is a work in progress. While Aria’s face is equipped with 16 micro motors that enable smiling and a range of facial expressions, the robot does not yet walk. It operates on a wheeled platform. This is a meaningful distinction from other humanoid robots in development by companies such as Agility Robotics, Figure AI, and Tesla, which are pursuing bipedal locomotion.
The design philosophy behind Aria is notable. Realbotix’s CEO, identified in the source as Kiguel, argues that looking human makes the robot approachable and helps avoid the “uncanny valley” effect—the phenomenon where almost-human robots repel rather than attract. The source quotes Kiguel citing World Health Organization figures indicating that close to a quarter of the world’s population suffers from loneliness. Kiguel’s stated position is that robots capable of engaging in social activities could reduce that suffering and potentially help people who are suicidal.
Third, the source places a specific valuation on the AI companionship market: $85 billion. This figure is attributed to Kiguel as an estimate. The source does not provide a breakdown of how that market is calculated, nor does it specify which segments are included. It is presented as a market-size estimate for the broader category of AI companionship, not solely for physical humanoid robots. European operators should note that this figure is an estimate from an industry executive, not an audited market study.
Fourth, the source provides anecdotal but compelling evidence of real-world adoption among older adults. The source describes Marvin Marcus, an 83-year-old resident who has formed a friendship with a robot named Meela. Marcus, a Yankees fan, uses a flip phone to call Meela and discuss baseball. His stated grievance is that the Yankees have not won a championship since 2009. He watches playoff games while discussing them with the bot, saying he cannot go into such detail with most other people but can blow off steam with Meela.
The source also quotes Skuler, presumably a company executive or industry figure, making a broader observation: the first humans who actually live with an AI and are building long-term relationships are not Silicon Valley geeks. They are older adults in the United States. This is a counterintuitive finding for many technology observers, but it aligns with the logic of the loneliness crisis. Older adults may have more time, more need for social interaction, and fewer alternatives for conversation partners.
Fifth, the source addresses the geographic and cultural factors driving humanoid robot adoption, with a specific focus on Japan. Coral Capital CEO James Riney, an investor in Japanese technology companies, is quoted identifying three factors. The first is a labor shortage and a desire to depend less on mass immigration. The second is a widespread cultural view of robots as friends—the source contrasts this as “more Doraemon vs. Terminator.” The third is Japan’s existing dominance in many parts of the robotics supply chain.
Finally, the source includes a cautionary note on adoption dynamics. A figure identified as Zhao is quoted saying that increased funding toward humanoid startups has accelerated the pace of progress, but that the most durable adoption comes from showing reliable and repeatable value in production or service operations, not just a one-off showcase. This is a critical point for European operators. Hype and venture capital can build prototypes, but sustained market penetration requires operational proof.
What it means for European operators
For European operators, the source material offers both opportunity and a warning. The opportunity is clear: the companion-robot market is not a distant future. It is happening now, with real products, real users, and real revenue potential. The warning is equally clear: the industry is young, the data is incomplete, and many early deployments may not survive contact with commercial reality.
European care facilities and senior-living operators should pay close attention to the anecdotal evidence from the United States. The source describes older adults forming genuine, long-term relationships with AI companions. This is not a hypothetical use case. It is a lived experience for at least some residents. For European operators facing their own demographic pressures—aging populations, stretched care budgets, and a shortage of qualified caregivers—companion robots could offer a partial solution. They are not a replacement for human care, but they can supplement it, providing conversation, engagement, and a degree of emotional support.
However, European operators must also be realistic about the current state of the technology. The source explicitly notes that Realbotix’s Aria, despite its hyper-realistic face, cannot walk. It is a wheeled platform. This limits its utility in environments with stairs, uneven floors, or other accessibility barriers. European care facilities, many of which are housed in older buildings, will need to assess whether such platforms can navigate their physical spaces. The source does not provide specifications on battery life, charging time, or maintenance requirements, so operators should not assume these are solved problems.
The $85 billion AI companionship market estimate is a headline number, but it is not a business plan. European operators should treat it as a signal of investor interest and consumer demand, not as a guaranteed revenue pool. The source does not break down this figure by geography, product type, or customer segment. It is entirely possible that the market is concentrated in specific regions or demographics that do not align with a given European operator’s target audience.
The Japanese experience, as described in the source, offers a useful comparative lens. Riney’s three factors—labor shortage, cultural acceptance of robots as friends, and supply-chain dominance—are not all present in Europe. European labor markets vary significantly by country. Some face acute shortages in specific sectors; others have high unemployment. The cultural view of robots in Europe is not monolithic. While some European countries may be receptive to robots as companions, others may be more skeptical. The source does not provide European-specific data, so operators should not assume that the Japanese model transfers directly.
The cautionary note from Zhao about “reliable and repeatable value” is perhaps the most important takeaway for European operators. The source does not provide specific examples of what constitutes reliable and repeatable value in a companion-robot context. It does not state whether Meela or Aria has been deployed in a commercial care setting, nor does it provide data on user satisfaction, retention, or clinical outcomes. European operators considering a pilot deployment should design their own metrics and evaluation criteria. They should not rely solely on vendor claims or media coverage.
The source also raises a question it does not answer: how many of the 13,317 humanoid robot shipments are commercial sales versus demos or pilots? This is a material gap in the data. European operators negotiating with robot vendors should ask for this breakdown directly. They should also ask about spare-part availability, software update policies, and the long-term viability of the vendor. The source does not provide any information on these operational details, so operators must gather it themselves.
There is also a regulatory dimension that the source does not address. European operators will need to consider data privacy, particularly if companion robots are deployed in care settings where they may record conversations or video. The source mentions that the app associated with one of the robots has customization features, including a custom word dictionary, audio clip retention settings, and voice learning. It also notes that the robot can learn to pick a user’s voice out of a group conversation and can be taught other voices it will commonly hear. These features have clear privacy implications. The source does not state whether the robot’s data is stored locally or in the cloud, nor does it specify retention periods. European operators must assume that any device with microphones and cameras will be subject to GDPR and other local regulations.
The source also does not address cybersecurity. A companion robot with network connectivity is a potential attack surface. European operators should not assume that the vendor has implemented robust security measures. They should request documentation, conduct their own risk assessments, and consider isolating these devices on separate network segments.
For European robotics startups, the source material offers a strategic insight: the market is bifurcating. There are humanoid robots designed for physical labor, such as those from Agility Robotics, Figure AI, and Tesla, and there are humanoid robots designed for social interaction, such as Realbotix’s Aria. These are different products with different engineering challenges, different go-to-market strategies, and different regulatory considerations. A startup that tries to do both may end up doing neither well. The source does not provide guidance on which path is more viable, but it does suggest that both paths have credible backers and stated market demand.
The source also highlights the importance of cultural context. Riney’s observation about Japan’s “Doraemon vs. Terminator” cultural view is specific to Japan. European startups should not assume that the same cultural acceptance exists in their home markets. They may need to invest in public education and trust-building, particularly if they are targeting consumer or care applications.
Finally, European operators should note the source’s acknowledgment that the industry is early-stage. The 13,317 units shipped last year is a tiny base. Many vendors may not survive to 2035. The 2.6 million unit projection assumes a near-doubling of shipments annually, which is an aggressive assumption. Operators should build flexibility into their contracts and avoid long-term lock-in with vendors whose financial stability is unproven.
The source does not provide any information on pricing, total cost of ownership, or return on investment for companion robots. European operators will need to develop their own financial models. They should also consider the human dimension: the source describes older adults forming genuine relationships with these robots. That is a positive outcome, but it also raises questions about dependency, emotional attachment, and what happens if the robot breaks or the vendor goes out of business. The source does not address these issues.
In summary, the source material describes a market in transition. Humanoid robots are no longer just workers. They are becoming friends, companions, and confidants. The market is projected to grow substantially, but the current base is small, and the data is incomplete. European operators should approach this space with optimism, but also with rigor. They should demand evidence, ask hard questions, and design their own evaluation frameworks. The technology is promising. The business case is still being written.
Sources
https://www.forbes.com/sites/johnkoetsier/2025/01/20/humanoid-robots-friends-not-just-workers/
Published by Vigla Media OÜ (Estonia).