Even as costs fall, the physical components—actuators, sensors, batteries, and the manufacturing capacity to produce them at scale—remain central constraints. Actuators alone account for roughly 50% of humanoid production costs, and 2026 production economics still sit in the $30,000 to $150,000 range per unit depending on capability and volume . Mass-market home adoption requires not just cheaper components, but entirely new supply chains.
General-purpose robots must navigate the unpredictable chaos of real homes, where performance depends on embodied interaction, dexterous manipulation, and reliability across infinite variables. Unlike large language models trained on internet text, physical AI lacks a comparable training corpus. The industry has not yet had its household-robot "GPT moment" .
Barclays frames humanoid robots as part of a much broader "physical AI" shift encompassing autonomous vehicles and drones—all of which compete for the same scarce computing resources, AI talent, and investment capital . This competition slows the focused R&D needed to solve household-specific challenges.
Near-term opportunity, Barclays argues, is concentrated in narrow-task robots operating in controlled environments like warehouses and welding stations—not in a general-purpose home robot that can fold laundry, cook dinner, and watch the kids .
The cautious near-term posture does not contradict Barclays' earlier projection of a $200 billion humanoid robotics market by 2035 . It is a matter of what is being measured and on what timeline.
In short, the analysts can see a massive industrial and specialized robot market by 2035 while remaining cautious about when general-purpose home robots will actually arrive.
The distance between today's most impressive demonstrations and a real consumer product is still substantial.
Humanoid robot costs have fallen from roughly $1 million-plus research platforms in 2020 to sub-$100,000 commercial units in 2026, with some Chinese models like Unitree's G1 priced as low as $16,000 for a base model . However, production economics still range from $30,000 to $150,000 per unit depending on capability and volume, and enterprise-grade systems from Boston Dynamics and Agility Robotics can exceed $250,000
. These price points are compatible with industrial deployment but remain far above mass-home-adoption levels.
Unitree's financials highlight the complexity. The company reported explosive growth in 2025—revenue of 1.7 billion yuan with a 60% gross margin on humanoid robots—and became the first profitable humanoid robotics company . Yet its Q1 2026 net profit fell 47.7% year over year as R&D and capacity spending surged
. The majority of its humanoid revenue still comes from enterprise reception, tour-guide, and research applications, not consumer homes
.
In January 2026, Hyundai Motor Group appointed Milan Kovac, the former head of Tesla's Optimus humanoid robot program, as an adviser and outside director of Boston Dynamics . Kovac had led the Optimus program before leaving Tesla in June 2025. His move to Boston Dynamics signals aggressive competition for the technical leadership needed to build scalable, viable products. It also underscores a broader industry reality: the race is still about R&D and product development, not about shipping millions of general-purpose home robots
.
Humanoid robots are making rapid, genuine progress—but almost all of it is happening in the controlled, narrow-task environments of factories, warehouses, and research labs. The same forces that make a $200 billion industrial market plausible by 2035 are precisely what keep general-purpose household deployment a decade or more away: until safety, hardware economics, data, and computing challenges are solved, the home robot that captured investor imagination will remain on the drawing board .
Comments
0 comments