The most recent coverage shows that humanoid robots are moving from prototype hype toward modest commercial roll‑out, but significant technical and economic hurdles remain. A McKinsey report cited in January 2026 notes that public demos still fall short of delivering reliable, cost‑effective performance, citing safety, limited battery life, dexterity and price as key barriers, and warns that even if Tesla reaches its goal of a million Optimus units by the end of 2027, the robots must prove they can work alongside humans for less than a human wage. Meanwhile, global installations of humanoid robots hit 16,000 units in 2025, with Chinese firms leading the market: Shanghai‑based AgiBot topped the list with over 5,000 units and a 31 % share, Unitree held 27 % and is pushing lower‑cost platforms, while UBTech, Leju and Tesla each hold roughly 5 % of installations. New business models such as robot‑as‑a‑service are gaining traction, especially for entertainment and retail deployments. At CES 2026, the British startup Humanoid displayed its HMND 01 biped, reporting about 25,000 pre‑orders and pilot programs with six Fortune 500 firms, while Boston Dynamics demonstrated its Atlas robot operating on Hyundai’s Georgia plant, underscoring a “sea change” in how the industry talks about physical AI. In parallel, Chinese company PNDbotics released a video of its Adam‑U Ultra robot performing a complex dance, showcasing advanced vision and lidar integration. Nonetheless, reliability concerns persist: Business Insider reported that XPeng’s IRON humanoid fell face‑first during its first public showcase in southern China, and Gartner’s 2026 outlook predicts fewer than 20 companies will be able to deploy humanoids at scale by 2028, describing the technology as still immature and not yet cost‑effective. Overall, the sector is seeing accelerated installations and high‑profile demos, but widespread, economically viable deployment remains limited.

Mentioned in this article