The Year Robots Stopped Performing and Started Working

For years, humanoid robots have been the darlings of tech conferences and viral videos, dazzling audiences with backflips, dance routines, and carefully choreographed demonstrations. But October 2026 may be remembered as the month the robotics industry finally walked off the demo stage and onto the factory floor. A wave of funding, commercial deployments, and shifted timelines from industry leaders suggests we are witnessing a genuine transition from experimental prototypes to revenue-generating machines.

Foundry Robotics Raises $100 Million for Autonomous Assembly

On October 28, 2026, Foundry Robotics announced an oversubscribed $100 million Series A funding round led by Marlinspike Partners, with participation from Khosla Ventures, Red Glass Ventures, SE Ventures, Wing VC, and others. The Emeryville, California-based startup is building what it calls America’s first autonomous factory for assembly, combining AI-powered robotic cells with experienced manufacturing teams to deliver manufacturing-as-a-service.

The company has expanded its manufacturing footprint twentyfold to a 130,000-square-foot facility in Emeryville, already shipping products to customers in defense, energy, and commercial sectors. Founder and CEO Adarsh Kulkarni, who previously served as Scale AI’s inaugural Head of Robotics, believes the future of American manufacturing requires starting from scratch rather than retrofitting existing factories.

Key details of the Foundry Robotics raise:

• $100 million Series A, led by Marlinspike Partners
• 130,000-square-foot autonomous factory in Emeryville, California
• Targets high-value, complex electro-mechanical products in defense and industrial sectors
• Manufacturing-as-a-service model eliminates need for customers to build dedicated production facilities

GXO Logistics Expands Humanoid Robot Deployment

Two days earlier, on October 26, GXO Logistics and Agility Robotics announced an expansion of their pioneering partnership, deploying Agility’s Digit humanoid robot to additional warehouse sites in North Carolina and Tennessee. This builds on their 2024 agreement that made GXO the first company to commercially deploy humanoid robots in live warehouse operations at Flowery Branch, Georgia.

The Digit robot has already moved more than 100,000 totes at the Georgia facility, working alongside existing automation to place containers onto conveyors. The expanded deployment will begin with Digit 4 series robots in early 2027, followed by Digit 5 units, which Agility unveiled on September 15, 2026. Digit 5 is designed for cooperatively safe work in close proximity to humans and addresses a broader range of industrial workflows.

Peggy Johnson, CEO of Agility Robotics, called the expansion the clearest signal yet that the commercial value of humanoid robots is real. The deployment uses Agility Arc, a cloud-based fleet management platform that integrates the robots with existing warehouse systems. GXO’s commitment to expanding beyond a single pilot site demonstrates that robotics-as-a-service models are gaining traction in logistics operations.

Unitree Founder Accelerates Humanoid Timeline to 2028

Perhaps the most striking shift came from Wang Xingxing, founder of Unitree Robotics, China’s largest humanoid robot manufacturer. Speaking at a physical AI conference in Hangzhou on October 23, Wang dramatically revised his timeline for humanoid robot breakthroughs. Just two months earlier at the World Robot Conference in Beijing, he estimated it would take ten years for humanoids to reach mass deployment. Now, he predicts robots will complete 80 percent of assigned tasks from verbal commands in factories and homes as soon as 2028.

Wang credits this accelerated timeline to rapid advances in AI models, particularly OpenAI’s GPT-6, which can now guide robots through previously untrained physical tasks. Unitree’s humanoids running GPT-6 can already paint pictures, changing colors and rinsing the palette, and play Go by precisely placing pieces on a game board. Wang defines the 80-percent task completion threshold as the aha moment for practical robotic applications, the point where humanoids transition from novelties to genuinely useful tools.

Unitree is preparing for massive scale. The company currently produces tens of thousands of robots annually but plans to expand capacity to meet potential demand for tens of millions of machines. This production scaling would also drive significant demand for AI chips, Wang noted. The company is developing its own AI models while collaborating with labs in China and abroad.

The Unitree Rehearsal System

Central to Unitree’s acceleration is a project Wang calls Unitree Rehearsal, which he named the company’s most important initiative of the year. The system uses AI to autonomously search research papers, write code, and test it in simulation before deploying on real robots. Feedback from physical trials is then collected to refine the models further. This self-improving loop dramatically reduces the time and cost of training robots for new tasks, effectively allowing humanoids to teach themselves new skills.

The H2-X Kung Fu Legend Guardian

Unitree also captured public attention on October 21 with the release of a one-minute video showcasing the H2-X, dubbed the Kung Fu Legend Guardian. The humanoid robot demonstrates remarkable physical capabilities, including punching and kneeing car doors in a scrapyard, dropping from a four-meter platform and rolling to absorb the impact, confronting a knife-wielding attacker to protect a person, and lifting a car to free a trapped individual.

While the theatrical demonstration generated significant social media buzz, Unitree has not disclosed pricing, full specifications, or a launch date for the H2-X. The company’s standard H2 model starts at $29,900 and stands 1.82 meters tall, weighing approximately 70 kilograms. The H2-X appears to be a specialized variant focused on physical prowess and security applications, though questions remain about whether such a robot could clear consumer safety certifications for personal bodyguard use.

The Broader Investment Landscape

The momentum extends well beyond individual companies. According to PitchBook data, robotics and physical AI startups raised a record $33.4 billion globally in the first half of 2026 alone, already surpassing the $28.7 billion raised in all of 2025. The Bay Area remains the epicenter, with approximately 1,000 robotics and physical AI companies occupying more than 7.6 million square feet of commercial space, up from less than 500,000 square feet in 2020.

Industry analysts note that the strongest investment momentum is shifting toward the software and components that underpin autonomous systems. AI and autonomy platforms reached $7.7 billion in trailing twelve-month funding, while actuators, motion systems, and end effectors also posted sharp growth. The infrastructure beneath the robots, not just the machines themselves, is attracting serious capital.

Shipment Data Confirms the Shift

The numbers tell a compelling story. Vendors shipped approximately 25,000 humanoid robots in the first half of 2026, already exceeding the roughly 18,500 shipped in all of 2025. More significantly, the share of robots going to display and validation purposes, including research, education, and performances, dropped from 84 percent in 2025 to about 70 percent in the first half of 2026. Real-world work nearly doubled its share to approximately 30 percent of all shipments.

Challenges Remain on the Path to Scale

Despite the optimism, significant hurdles persist. Wang himself acknowledged that if you take a humanoid home today and ask it to clean the room, it honestly cannot complete that task. The gap between laboratory demonstrations and reliable real-world performance remains substantial. Regulatory challenges, including the FCC’s robotics restrictions in the United States, are reshaping supply chains and potentially redirecting market volume to Europe.

Battery life continues to be a limiting factor, with even Unitree’s H2-X offering approximately eight hours of standby endurance, likely much less during active operation. Safety certifications for robots working alongside humans require rigorous testing, and the economics of robotics-as-a-service models are still being validated as deployments expand beyond pilot programs.

Nevertheless, the convergence of three critical factors, improved AI models, massive capital investment, and proven commercial deployments, suggests the robotics industry has reached an inflection point. The question is no longer whether robots will work alongside humans, but how quickly and in what numbers. If the events of October 2026 are any indication, the answer may be sooner than anyone expected.

---

Edited by Palawan @QUE.COM
Website: https://QUE.COM Intelligence
Sponsored by: https://MAJ.COM AI Autonomous