Automate 2026 and the humanoid robotics reality check: China moves to mass production, the West still at the pilot stage
🔎 Why June 2026 marks a tipping point
Humanoid robotics has just crossed an invisible line. Not the one of democratization — we are far from that. The one of institutional denial: humanoids are no longer lab projects, they are entering assembly lines. Automate 2026, which opens its doors from June 22 to 25 at McCormick Place in Chicago, crystallizes this shift.
North America's most important industrial automation show is dedicating an entire Humanoid Robot Pavilion this year, sponsored by NVIDIA. Live commercial deployments from Boston Dynamics are expected there. Figure AI is presenting its automated production strategy there. The signal is clear: American industry is organizing to scale up.
Except that on the other side of the world, the shift has already happened. China is deploying humanoids in factories by the hundreds, its automakers are selling them online, and its public governance bodies (MIIT, SASAC) are pushing for accelerated industrialization. The gap between Western discourse and Chinese reality has never been wider.
The Essentials
- Automate 2026 (June 22–25, Chicago) marks the official transition from demo to production in the West, with a Humanoid Robot Pavilion sponsored by NVIDIA and live deployments from Boston Dynamics.
- Figure AI reaches 1 robot/hour in its BotQ factory (June 2026), with a projected capacity of 12,000 units/year — the first Western demonstration of sustained automated production throughput.
- China moves beyond the pilot stage: BYD integrates humanoids into production, Chery sells the Mornine M1 online (~€36,000), Seres unveils its Xiaosai robot, and XPeng merges autonomous AI and robotics.
- Europe seeks niches (harvesting, hospitality) in the face of Chinese giants, as shown at Vivatech 2026.
- Chinese control of rare earths directly impacts Boston Dynamics' ambitions and its planned IPO, according to DigiTimes.
Recommended Tools
| Actor | Model/Product | Stage (June 2026) | Ideal for |
|---|---|---|---|
| Figure AI | Figure (BotQ) | Automated production (1/hour) | Western mass production |
| Boston Dynamics | Atlas | Hyundai customer field testing | Heavy industrial handling |
| Seres | Xiaosai | Unveiling (June 2026) | Chinese auto integration |
| Chery | Mornine M1 | Online sales (~€36,000) | First Chinese consumer humanoid |
| NVIDIA | Jetson Platform / Pavilion | AI Infrastructure | Brain of humanoids |
China is no longer just doing pilots: it's producing
The fundamental difference in Q2 2026 is that China has stopped announcing "pilots" to move directly to industrial deployment. The June 2026 StartupsOfa report puts it bluntly: Chinese mass production now outpaces Western pilot deployments.
BYD, China's leading automaker, is already integrating humanoid robots into its production lines. The goal is not to test if it works — it is to reduce reliance on manual labor and automate handling tasks that traditional robots cannot manage. This is an industrial choice, not a PR exercise.
Chery took a further step in April 2026 by launching the online sale of its Mornine M1 robot, at around €36,000. A price that is far from incidental: it positions the humanoid as affordable professional equipment, not as a research prototype. XPeng, for its part, is explicitly linking its autonomous driving systems to its robotics branch, sharing Perception AI.
Seres (SH: 601127) joined the movement on June 15, 2026, by unveiling its "Xiaosai" robot, as reported by CnEVPost. Director and VP Kang Bo released a demonstration video. Seres becomes the latest in a long list of Chinese automakers to enter the humanoid space — an auto-robotic convergence that exists at this scale nowhere else.
This convergence is not spontaneous. The MIIT (Ministry of Industry and Information Technology) and the SASAC (State-owned Assets Supervision and Administration Commission) are guiding and financing this transition. Star Robotics and other players benefit from a public-private ecosystem with no equivalent in the West.
Figure AI and BotQ: the Western exception that proves the rule
Faced with the Chinese machine, the West holds one trump card: Figure AI and its BotQ factory.
In June 2026, Figure AI announces that BotQ has reached a rate of 1 robot/hour. According to Axis Intelligence, this is the first time a humanoid manufacturer has demonstrated a sustained automated production throughput. The first generation of the BotQ line is sized to produce up to 12,000 robots per year, according to The Robot Report.
The detail that changes everything: the Figure robots themselves are used in the manufacturing process to build other humanoids. It is a self-amplifying loop. Figure AI projects that the number of robots involved on the line will gradually increase to automate more and more steps.
It is the only Western factory that resembles, even from afar, what China is deploying. But 12,000 units/year remains an order of magnitude below Chinese industrial capacities when aggregating the volumes of BYD, Chery, Seres and the others. Figure AI proves that the West knows how to do it. But proving is not producing at the scale of a country.
Boston Dynamics Atlas : le robot humanoïde qui fait tout seul — Atlas field testing at Hyundai
Boston Dynamics is no longer a startup churning out viral demos. The acquisition by Hyundai has changed the nature of the game.
Atlas is now officially positioned as an "enterprise humanoid robot designed for real industrial work, material handling, and intelligent automation." The Boston Dynamics product page no longer talks about parkour or backflips: it talks about task sequencing in customer environments.
In practice, Atlas is taking its first steps in a Hyundai facility for field testing on real task sequencing operations. This is the critical phase — where the rubber meets the industrial asphalt. A robot that walks in a lab is worthless. A robot that sorts parts for 8 hours in a Hyundai factory is a product.
The projet Atlas remains, however, in the field testing phase, not mass deployment. The ambition is there — but the deployment pace remains calibrated to industrial pilots, not hundreds of units rolling out every month.
Automate 2026: the trade show that makes the shift official
Automate 2026 is at Chicago's McCormick Place from June 22 to 25. The largest automation trade show in North America. This year, it plays a different role than in its previous editions: it serves as an institutional springboard for the industrial humanoid.
The Humanoid Robot Pavilion, sponsored by NVIDIA, is the physical heart of this shift. TechTimes reports that Figure AI is presenting its BotQ factory and its throughput of 1 robot/hour there. Boston Dynamics is doing live demonstrations of commercial deployments there. Robotics 24/7 points out that the Humanoid Robot Forum accompanies the pavilion, with a focus on vision systems and embedded AI.
The signal is not technological — it is economic. When NVIDIA sponsors an entire pavilion dedicated to humanoids, it means the value chain (chips, simulators, cloud infra) is ready to monetize. When deployments happen "live" rather than in pre-recorded video, it means integration teams have reproducible results to show.
Automate 2026 does not create the trend. It sanctifies it.
Rare earths: the Achilles' heel of Western robotics
All this momentum runs up against a geopolitical wall that few commentators want to see. DigiTimes, in a May 19, 2026 article, reveals that Chinese control of rare earths casts a "rare shadow" over Atlas's ambitions and Boston Dynamics' planned IPO.
Rare earths are not a minor detail. High-density motors, the permanent magnets in actuators, advanced sensors — all of these depend on materials for which China controls over 60% of the mining and nearly 90% of the refining. A humanoid like Atlas consumes dozens of kilograms of them per unit.
For Hyundai and Boston Dynamics, the equation is simple: wanting to produce 30,000 units/year by 2028 without controlling the upstream mining is like building a factory with no guarantee of supply. The analysts cited by DigiTimes are unequivocal: this dependence weighs on Boston Dynamics' valuation and the credibility of its volume projections.
China, ironically, does not face this problem. Its manufacturers (BYD, Chery, Seres) rely on a nationally integrated, vertically aligned supply chain. The Chinese competitive advantage in humanoid robotics is not just industrial — it is mineral and geopolitical.
Unitree G1 at Haneda Airport: Chinese humanoid robots make their way to Japan — physical robotics scales up — The export of Chinese robotics
Chinese mass production is not confined within the country's borders. The example of the Unitree G1 deployed at Haneda Airport in Japan illustrates a game-changing export dynamic.
A Chinese humanoid robot in a Japanese airport—a country renowned for its own robotic ecosystem—is a powerful symbol. It means that the value for money of Chinese humanoids now allows them to compete with local players on their own turf. Physical robotics is scaling up, and that scale is Chinese.
Japanese and Korean manufacturers, historically dominant in conventional industrial robotics, are now facing a wave of Chinese humanoids sold at a fraction of the price. The G1 is not a luxury robot—it is functional equipment deployed in a highly demanding operational environment.
Genesis AI unveils GENE-26.5 and humanoid robotic hands: robotics goes full-stack — Full-stack innovation as a response to mass production
Faced with the Chinese advantage in volume, some players are betting on cutting-edge technological differentiation. Genesis AI and its GENE-26.5 with humanoid robotic hands embody this strategy: rather than competing on unit price, they are pushing the boundaries of dexterity.
The robotic hand is the most complex component of a humanoid. An arm that lifts 20 kg is mechanics. A hand that grasps an egg without breaking it and then grips a tool is real-time sensor-actuator-AI integration. Genesis AI is betting on this complexity as a barrier to entry.
This full-stack approach makes sense if markets value fine manipulation capability more than unit cost. But faced with hundreds of Mornine M1s deployed in Chinese factories at €36,000, the question is: how many customers are willing to pay a significant premium for superior dexterity? The answer will determine whether Western technological innovation can compensate for the Chinese industrial disadvantage.
Europe's niches: grape harvesting, reception, and Vivatech 2026
Europe is not playing in the same league as China or the United States. And it knows it.
TechXplore reports that at Vivatech 2026 in France, humanoids capable of harvesting grapes or welcoming visitors were in the spotlight. European startups explicitly sought to occupy niches there beyond what the Chinese giants offer. Robotic harvesting, hospital reception, agricultural logistics — these are specialty markets where volume is not the primary criterion.
It is a pragmatic strategy, but one that has its limits. Niches do not create economies of scale. Without volume, component costs remain high. Without competitive costs, exporting becomes difficult. And without exports, the European humanoid industry remains an ecosystem of subsidy-funded startups, not an industry.
The risk for Europe is to find itself in the situation it already knows in the cloud or semiconductors: excellent in R&D, marginal in production. Except that physical robotics differs from software in one key way: you cannot "deploy" a robot with a git push. You have to manufacture it, deliver it, maintain it.
❌ Common mistakes
Mistake 1: Confusing pilot and production
A pilot in a Hyundai plant or a deployment of 5 robots at a client's site does not constitute mass production. The distinction is fundamental: a pilot proves a concept, production proves a business model. Many Western commentators treat the two as equivalent, which distorts the perception of the balance of power.
Mistake 2: Ignoring the supply chain
Talking about "30,000 units/year" targets without mentioning the dependence on Chinese rare earths is writing industrial fiction. The DigiTimes report is clear: this is not a distant risk, it is a current brake on Boston Dynamics' IPO. Any analysis that bypasses this point is incomplete.
Mistake 3: Comparing prices without comparing contexts
The Mornine M1 at ~€36,000 and an Atlas, whose price is not public, are not in the same category. But presenting the Chinese price as "too low to be credible" without analyzing public subsidies (MIIT, SASAC), national economies of scale, and vertical integration demonstrates a misunderstanding of the Chinese cost structure.
Mistake 4: Overestimating the impact of Automate 2026
Automate 2026 is a symbolic tipping point for the West, not an event that changes the geopolitics of robotics. The Humanoid Robot Pavilion is important for the American ecosystem. It does not change the fact that China already produces more than all Western players combined.
❓ Frequently Asked Questions
What is Figure AI's production rate in June 2026?
Figure AI announces 1 robot/hour in its BotQ factory, representing a theoretical capacity of 12,000 units/year for the first generation of the line. This is the first Western demonstration of a sustained automated production throughput for a humanoid (Axis Intelligence, June 2026).
What is the price of Chery's Mornine M1 robot?
The Mornine M1 has been sold online since April 2026 for approximately €36,000. To date, it is the most openly marketed Chinese humanoid aimed at the professional general public (Les Numériques, 2026).
What role does NVIDIA play at Automate 2026?
NVIDIA sponsors the Humanoid Robot Pavilion at Automate 2026 in Chicago. Its role is to establish the Jetson platform and AI simulation tools as the reference infrastructure for Western humanoid developers (Robotics 24/7, June 2026).
Are rare earths really blocking Boston Dynamics?
Yes, partially. DigiTimes (May 2026) reports that Chinese control of rare earths casts a shadow over the production ambitions for Atlas and the Boston Dynamics IPO. High-density actuators depend on them, and no Western alternative is yet mature on an industrial scale.
Where does Europe stand in humanoid robotics?
Europe is positioning itself in niche markets (agriculture, hospitality, healthcare) as shown at Vivatech 2026. TechXplore (June 2026) notes that European firms are seeking out markets that Chinese giants are not yet targeting, but without significant production volumes.
✅ Conclusion
Humanoid robotics in Q2 2026 resembles a race where the competitors are not running on the same track. China produces, deploys, and exports. The West — with the notable exception of Figure AI and BotQ — is still trying to prove that its robots can hold up in real-world conditions. Automate 2026 makes this realization official, but the real reality check is geopolitical: without an autonomous rare earth supply chain, Western production ambitions will remain just ambitions.