Humanoid Robot Production 2030 Explained: J.P. Morgan Sees 400,000 a Year, a $100,000 Saving and Four Bottlenecks

Image: Tesla's Optimus humanoid at the October 2024 Robotaxi event in California. The production line J.P. Morgan's forecast depends on is a newer story — this is the robot at a public demonstration. Image: Steve Jurvetson, CC BY 2.0, via Wikimedia Commons.

TL;DR: Humanoid robot production 2030 explained: J.P. Morgan expects the US to build 300,000–400,000 humanoid robots a year by 2030, rising to about 1.5 million by 2035. The forecast is a bet that robots are moving out of demos and onto factory floors — but hands, actuators, the AI "brain" and production capacity still limit how fast that can happen.

Key Highlights

• 300,000–400,000 humanoid robots a year by 2030 in the US, and about 1.5 million by 2035, according to J.P. Morgan's forecast published on 12 November 2026.
• Costs are nearing parity with human labour, the bank says — and running a robot across multiple shifts could save a customer about $100,000 a year.
• Four bottlenecks cap the pace: hands, actuators, the AI "brain" and production capacity. Most US makers outside Tesla aim for just 10,000–30,000 units a year.
• The parts matter more than the robots right now: J.P. Morgan expects the market for actuators, motors and vision systems to approach $30 billion by 2035.
• Five names to watch: Tesla's Optimus, Agility Robotics' Digit, Boston Dynamics' Atlas, Figure AI's Figure robots and Apptronik's Apollo.

Humanoid robot production 2030 explained: what J.P. Morgan actually forecast

On Thursday 12 November 2026, J.P. Morgan put a number on a shift that has been visible all year: humanoid robots are leaving the demonstration stage and starting real work. The bank's forecast, reported by Reuters, is that production of humanoid robots in the United States will reach 300,000–400,000 units a year by 2030, and about 1.5 million units by 2035.

Three forces are driving it, according to the bank: labour shortages, improving economics and rising production capacity. Factories, warehouses and logistics firms are looking to automate repetitive tasks that are hard to staff, and the robots are finally cheap and capable enough to be worth trying at scale.

Politics is pushing the same way. The Trump administration has found that imported robotics and industrial machinery pose national security risks — a finding that could lead to tariffs or other measures designed to boost domestic production. That would favour US-based makers and the suppliers that feed them.

One thing to keep in mind throughout this post: a forecast is not a fact. J.P. Morgan is estimating what the industry could build if today's constraints ease. The 2030 number is four years of scaling away, and the bank itself says progress is still supply-constrained.

The numbers behind the forecast

Figure	Value	What it means
US humanoid production, 2030	300,000–400,000 units a year	J.P. Morgan's forecast for annual output, up from demonstration-scale volumes today
US humanoid production, 2035	About 1.5 million units a year	Roughly four times the 2030 figure, if scaling continues
Annual target, most US makers (except Tesla)	10,000–30,000 units	How far today's plans are from hundreds of thousands — the gap the forecast must close
Saving from multi-shift deployment	About $100,000 per customer a year	A robot working around the clock where a person works one shift
Key-components market, 2035	Approaching $30 billion	Actuators, motors and vision systems — the parts every humanoid needs

Image: Tesla's Optimus v.2 on display in a showroom in November 2025. Tesla said in its third-quarter update in October 2026 that Optimus production had begun on high-volume lines at its Fremont factory — though it disclosed no robot count. Image: Sikander Iqbal, CC BY-SA, via Wikimedia Commons.

Why factories want humanoids now

The pitch is simple: a humanoid fits a workplace built for humans. Unlike a fixed robotic arm or a wheeled cart, it can walk the same aisles, climb the same steps and handle the same tools and totes as a person — no factory redesign required. That is why the first deployments are in warehouses and on production lines, moving boxes, kitting parts and tending machines.

There is already evidence the shift is real, not just a forecast. GXO's plan to put 1,000 Apollo robots in its warehouses from late 2027 is the largest deployment announced so far, and Agility's Digit is already working in Amazon's warehouses. Tesla, meanwhile, said in its third-quarter update in October 2026 that Optimus production had begun on high-volume lines at its Fremont factory — the first production-quality robots it has built, though it gave no numbers on how many.

The economics are what turn pilots into purchase orders. J.P. Morgan estimates the cost of deploying a humanoid is approaching the cost of employing a human worker — and because a robot can work across shifts without tiring, the bank puts the saving at about $100,000 a year per customer. That is the number factory managers will actually run their budgets against.

The four bottlenecks holding robots back

J.P. Morgan is blunt about what is still in the way: "Near-term progress remains supply constrained, with a few binding bottlenecks: hands, actuators, the brain, and production capacity." Each one is a real engineering problem.

Hands have to grip reliably — picking up different shapes, weights and textures without dropping or crushing them. Actuators — the motors and joints that move the limbs — must be strong and small enough to fit a human-sized body. The brain is the AI that plans movement and learns tasks; NVIDIA's GR00T 2 robot brain, which learns dexterous skills from human video, is one attempt at solving it. And production capacity is the dullest bottleneck of all: most US makers outside Tesla are targeting only 10,000–30,000 units a year, a long way from hundreds of thousands.

That is why J.P. Morgan thinks the money, in the near term, is in the parts rather than the finished robots. It expects the market for key components — actuators, motors and vision systems — to approach $30 billion by 2035, and it favours suppliers such as Timken, Robotis, Leader Drive, Orbbec and Cognex. Over time, the bank expects value to shift towards the robot makers and platform integrators as the industry matures.

Why it matters for India readers

J.P. Morgan's forecast is about the United States, and it gave no India-specific production number — so do not read an Indian forecast into it. But the story still matters here, for three reasons.

First, India's own robotics industry is moving in the same direction: Peppermint Robotics' ₹100 crore funding shows Indian capital flowing into warehouse and factory robots. Second, the components market is global — Indian electronics manufacturers that can make actuators, motors and sensors stand to gain from a $30 billion parts market. And third, the tariff angle cuts both ways: if the US taxes imported robotics, Indian robotics firms hoping to sell into the American market would face the same barrier as everyone else.

One honest caveat: the "costs nearing parity with labour" maths lands very differently in India, where wages are far lower than in the US. Humanoid deployment in Indian factories is likely to lag the US and to start with hazardous or hard-to-staff tasks rather than as a straight replacement for workers.

Should you care?

✅ Yes, if you follow robotics, manufacturing or Tesla. A major bank putting 400,000 robots a year by 2030 on paper is the clearest sign yet that humanoids are becoming an industrial product, not a demo. ⏳ But keep the forecast in perspective: it is J.P. Morgan's estimate, the bottlenecks are real, and most makers are still planning in the tens of thousands. Watch the production numbers — especially Tesla's — over the next year, not the headlines.

FAQs

How many humanoid robots will be produced by 2030?
J.P. Morgan forecasts US production of 300,000–400,000 humanoid robots a year by 2030, rising to about 1.5 million by 2035. That is the bank's estimate for annual output, not a count of robots already built — and most US makers outside Tesla are currently targeting only 10,000–30,000 units a year.

Which companies make humanoid robots?
J.P. Morgan names five leading US makers: Tesla with Optimus, Agility Robotics with Digit, Boston Dynamics with Atlas, Figure AI with its Figure robots, and Apptronik with Apollo. Agility's Digit is already working in Amazon warehouses, and GXO plans to deploy up to 1,000 Apollo robots from late 2027.

How much does a humanoid robot cost?
J.P. Morgan did not publish a unit price. What it did say is that the cost of deploying a humanoid is nearing parity with the cost of employing a human worker, and that running a robot across multiple shifts could save a customer about $100,000 a year.

What is stopping humanoid robots from being deployed faster?
Four bottlenecks, in J.P. Morgan's words: hands, actuators, the brain and production capacity. More broadly, deployment is constrained by hardware, AI training data, computing power and production capacity — the supply chain for robot parts has to scale before the robots can.

Takeaway

J.P. Morgan's 400,000-a-year forecast says humanoids are crossing from demonstration to deployment — but the crossing depends on hands, actuators, brains and factories all scaling at once. If you want to see what deployment actually looks like, read our explainer on GXO's plan to put 1,000 Apollo robots in its warehouses.

Sources

CNBC TV18 — US humanoid robot production could hit 400,000 by 2030 as deployment ramps, JPMorgan says (12 Nov 2026, Reuters report)
The Standard — US humanoid robot production could hit 400,000 by 2030 as deployment ramps, J.P. Morgan says (12 Nov 2026, Reuters report)
TimesLIVE — US humanoid robot production could hit 400,000 by 2030: JP Morgan (12 Nov 2026, Reuters report)
DD India — US humanoid robot production could hit 400,000 by 2030 as deployment ramps, J.P. Morgan says (12 Nov 2026, Reuters report)
The Telegraph India — Humanoid robot production in US may hit 400,000 by 2030 as deployment ramps, says JP Morgan (12 Nov 2026, Reuters report)
TipRanks — Tesla Stock Dips Despite Being Hailed as Key Builder of Over a Million American Humanoids (12 Nov 2026)