<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>Biped News</title>
  <link href="https://biped.news/" />
  <link rel="self" href="https://biped.news/feed.xml" />
  <id>https://biped.news/</id>
  <updated>2026-09-15T18:08:43.802Z</updated>
  <subtitle>News and analysis covering the emerging humanoid robotics industry</subtitle>
  <author>
    <name>Biped News</name>
  </author>
  <entry>
    <title>PHYMI Raises Nearly $100 Million to Build Physical Agents Beyond the Lab</title>
    <link href="https://biped.news/article/phymi-physical-agent-seed-round-2026" />
    <id>biped-article-phymi-physical-agent-seed-round-2026</id>
    <updated>2026-09-13T02:00:00.000Z</updated>
    <published>2026-09-13T02:00:00.000Z</published>
    <summary>PHYMI, founded in 2026 by former DeepRoute.ai executive Nianqiu Liu, says it has closed a nearly $100 million seed round led by IDG Capital with Yunqi Capital, DiDi, Fosun RZ Capital, Yaotu Capital, and Hesai Technology participating. The funding is significant for embodied AI because PHYMI is targeting long-horizon movement and manipulation in open, dynamic environments, while still lacking public proof of a product, customer deployment, or commercial robot specs.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/phymi-physical-agent-seed-round-2026-hero.png" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="PHYMI" />
    <category term="Physical AI" />
    <category term="Robotics Funding" />
    <category term="China robotics" />
    <category term="Physical Agents" />
  </entry>
  <entry>
    <title>XPeng Opens IRON Production Line as Humanoid Race Turns to Manufacturing</title>
    <link href="https://biped.news/article/xpeng-iron-production-line-2026" />
    <id>biped-article-xpeng-iron-production-line-2026</id>
    <updated>2026-09-08T02:00:00.000Z</updated>
    <published>2026-09-08T02:00:00.000Z</published>
    <summary>XPeng has opened a production line for its IRON humanoid robot, according to fresh CnEVPost reporting based on the company announcement. The milestone matters because it shifts IRON from a mass-production promise toward manufacturing execution, while leaving key questions about capacity, price, yield, and external customers unanswered.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/xpeng-iron-production-line-2026-hero.webp" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="XPeng" />
    <category term="IRON" />
    <category term="Humanoid Robots" />
    <category term="robot manufacturing" />
    <category term="Physical AI" />
    <category term="China robotics" />
  </entry>
  <entry>
    <title>Humanoid Robot End Effectors Explained: Hands, Grippers, and Tooling for Real Work</title>
    <link href="https://biped.news/article/humanoid-robot-end-effectors-tooling-explained" />
    <id>biped-article-humanoid-robot-end-effectors-tooling-explained</id>
    <updated>2026-09-09T03:10:00.000Z</updated>
    <published>2026-09-08T01:00:00.000Z</published>
    <summary>Humanoid robot end effectors range from dexterous hands to rugged grippers and tool changers. The choice shapes payload, reliability, cost, and deployment speed.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/humanoid-robot-end-effectors-tooling-explained-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="End Effectors" />
    <category term="Robot Hands" />
    <category term="Manipulation" />
    <category term="Tool Changers" />
    <category term="Deployment" />
  </entry>
  <entry>
    <title>IFA 2026 Shows Humanoid Robots Moving From Show Tricks to Product Segments</title>
    <link href="https://biped.news/article/ifa-2026-humanoid-platforms-market-shift" />
    <id>biped-article-ifa-2026-humanoid-platforms-market-shift</id>
    <updated>2026-09-07T02:05:48.435Z</updated>
    <published>2026-09-07T02:00:00.000Z</published>
    <summary>Fresh IFA 2026 coverage shows humanoid robots spreading across distinct product categories rather than converging on one general-purpose machine. PrimeBOT Q1, Booster K1, EngineAI T800, Agibot A3, and Zero W1 point to a market where developer access, runtime, payload, sensing, and price matter more than stage performance.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/ifa-2026-humanoid-platforms-market-shift-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="IFA 2026" />
    <category term="humanoid robots" />
    <category term="physical AI" />
    <category term="Booster Robotics" />
    <category term="EngineAI" />
    <category term="Agibot" />
    <category term="PrimeBOT" />
    <category term="Zero" />
  </entry>
  <entry>
    <title>Skild S1 Puts Robot Training on the Single-Video Clock</title>
    <link href="https://biped.news/article/skild-s1-single-video-robot-training-2026" />
    <id>biped-article-skild-s1-single-video-robot-training-2026</id>
    <updated>2026-09-06T02:04:22.423Z</updated>
    <published>2026-09-06T02:00:00.000Z</published>
    <summary>Skild AI’s S1 model puts the spotlight on robot training speed. If single-video task learning proves reliable, it could lower deployment friction for humanoids and mobile manipulators, but production proof still needs measured success rates, recovery data, and customer evidence.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/skild-s1-single-video-robot-training-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Skild AI" />
    <category term="S1" />
    <category term="physical AI" />
    <category term="robot foundation models" />
    <category term="robot training" />
    <category term="humanoid robots" />
  </entry>
  <entry>
    <title>Nissan Turns Factory AMRs Into a Physical AI Reality Check</title>
    <link href="https://biped.news/article/nissan-smyrna-amr-physical-ai-2026" />
    <id>biped-article-nissan-smyrna-amr-physical-ai-2026</id>
    <updated>2026-09-05T02:05:02.684Z</updated>
    <published>2026-09-05T02:00:00.000Z</published>
    <summary>Nissan is rolling out OTTO Motors autonomous mobile robots at its Smyrna, Tennessee factory. Public reporting says the largest units can move up to 4,190 pounds and coordinate as a fleet inside the body shop, affecting 64 forklift and tug roles while Nissan says workers will be reassigned or retrained. The deployment is a useful physical AI benchmark for humanoid robotics because it shows what factory buyers already value: payload, fleet coordination, uptime, intervention rates, and measurable material flow.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/nissan-smyrna-amr-physical-ai-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Nissan" />
    <category term="Smyrna" />
    <category term="OTTO Motors" />
    <category term="Rockwell Automation" />
    <category term="autonomous mobile robots" />
    <category term="physical AI" />
    <category term="factory automation" />
    <category term="AMRs" />
    <category term="material handling" />
    <category term="robot deployment" />
  </entry>
  <entry>
    <title>UBS Says Physical AI&apos;s First Winners May Be the Parts Suppliers</title>
    <link href="https://biped.news/article/ubs-physical-ai-picks-shovels-supply-chain-2026" />
    <id>biped-article-ubs-physical-ai-picks-shovels-supply-chain-2026</id>
    <updated>2026-09-04T02:05:56.440Z</updated>
    <published>2026-09-04T02:00:00.000Z</published>
    <summary>UBS is pushing investors to look beyond individual humanoid robot makers and toward the enabling supply chain behind physical AI. A September 3 Business Insider report on a UBS client note highlighted three areas: actuators and precision gears, sensors and perception, and edge compute plus power infrastructure. The article explains why these layers may determine whether humanoid robotics can move from impressive demos into reliable industrial equipment.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/ubs-physical-ai-picks-shovels-supply-chain-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="UBS" />
    <category term="physical AI" />
    <category term="humanoid robots" />
    <category term="robotics supply chain" />
    <category term="actuators" />
    <category term="precision gears" />
    <category term="robot sensors" />
    <category term="edge AI" />
    <category term="power electronics" />
    <category term="robotics investing" />
  </entry>
  <entry>
    <title>Berkeley Humanoid Lite Makes the $5,000 Robot a Serious Research Tool</title>
    <link href="https://biped.news/article/berkeley-humanoid-lite-open-source-hardware-2026" />
    <id>biped-article-berkeley-humanoid-lite-open-source-hardware-2026</id>
    <updated>2026-09-01T02:04:48.877Z</updated>
    <published>2026-09-01T02:00:00.000Z</published>
    <summary>Berkeley Humanoid Lite is an open-source, sub-$5,000 humanoid research platform built around 3D-printed modular cycloidal gearboxes, public CAD, embedded code, and training and deployment frameworks. It is not a deployment robot, but it could make physical AI research more reproducible.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/berkeley-humanoid-lite-open-source-hardware-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Berkeley Humanoid Lite" />
    <category term="open-source robotics" />
    <category term="3D-printed robots" />
    <category term="physical AI" />
    <category term="cycloidal gearboxes" />
    <category term="humanoid research" />
    <category term="sim-to-real" />
  </entry>
  <entry>
    <title>Series Elastic Actuators Explained: Why Robot Joints Sometimes Need Springs</title>
    <link href="https://biped.news/article/series-elastic-actuators-humanoid-robots-explained" />
    <id>biped-article-series-elastic-actuators-humanoid-robots-explained</id>
    <updated>2026-09-01T01:09:56.449Z</updated>
    <published>2026-09-01T01:00:00.000Z</published>
    <summary>Series elastic actuators use measured spring deflection to estimate torque and protect robot joints, but they add control complexity and bandwidth limits.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/series-elastic-actuators-humanoid-robots-explained-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="series elastic actuators" />
    <category term="humanoid robots" />
    <category term="robot joints" />
    <category term="force control" />
    <category term="torque control" />
    <category term="compliance" />
    <category term="legged robots" />
  </entry>
  <entry>
    <title>Microduck Turns Biped Robot Learning Into a $399 Developer Platform</title>
    <link href="https://biped.news/article/microduck-open-source-biped-robot-2026" />
    <id>biped-article-microduck-open-source-biped-robot-2026</id>
    <updated>2026-08-31T02:04:40.828Z</updated>
    <published>2026-08-31T02:00:00.000Z</published>
    <summary>Microduck is a 25 cm, $399 open-source biped from Hugging Face and Pollen Robotics. It is not a practical home robot, but it could matter as a low-cost developer platform for robot learning, sim-to-real training, fall recovery, and community-shared motion policies.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/microduck-open-source-biped-robot-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Microduck" />
    <category term="Hugging Face" />
    <category term="Pollen Robotics" />
    <category term="open-source robotics" />
    <category term="physical AI" />
    <category term="reinforcement learning" />
    <category term="biped robots" />
  </entry>
  <entry>
    <title>South Korea&apos;s $1.66B Humanoid Plan Turns Physical AI Into Industrial Policy</title>
    <link href="https://biped.news/article/south-korea-humanoid-robotics-roadmap-2030" />
    <id>biped-article-south-korea-humanoid-robotics-roadmap-2030</id>
    <updated>2026-08-30T02:05:40.159Z</updated>
    <published>2026-08-30T02:00:00.000Z</published>
    <summary>South Korea is treating humanoid robots as an industrial policy and supply-chain sovereignty issue, not only a research category. The plan targets 80% domestic component localization by 2030, public procurement of 1,080 humanoids, a data cleanroom, and a domestic robot OS by 2031.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/south-korea-humanoid-robotics-roadmap-2030-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="South Korea" />
    <category term="humanoid robots" />
    <category term="physical AI" />
    <category term="industrial policy" />
    <category term="robotics supply chain" />
  </entry>
  <entry>
    <title>Meta&apos;s Data Center Robot Tests Move Physical AI Into the Server Aisle</title>
    <link href="https://biped.news/article/meta-data-center-robots-physical-ai-2026" />
    <id>biped-article-meta-data-center-robots-physical-ai-2026</id>
    <updated>2026-08-29T02:05:23.410Z</updated>
    <published>2026-08-29T02:00:00.000Z</published>
    <summary>WIRED reported on August 28 that Meta is testing robots from vendors including Kinova, ABB, and Watney Robotics inside data centers. The work includes cable handling, power cycling, inventory checks, rack transport, and part reseating. Biped.News examines why this matters for physical AI, what is actually confirmed, where the technical limits remain, and why data centers are becoming a serious robotics proving ground.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/meta-data-center-robots-physical-ai-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Meta" />
    <category term="data centers" />
    <category term="physical AI" />
    <category term="robotics" />
    <category term="ABB" />
    <category term="Kinova" />
    <category term="Watney Robotics" />
    <category term="automation" />
    <category term="infrastructure" />
  </entry>
  <entry>
    <title>Renesas Opens Beijing Physical AI Lab as Humanoid BOM Race Gets Serious</title>
    <link href="https://biped.news/article/renesas-physical-ai-lab-beijing-2026" />
    <id>biped-article-renesas-physical-ai-lab-beijing-2026</id>
    <updated>2026-08-28T02:05:17.953Z</updated>
    <published>2026-08-28T02:00:00.000Z</published>
    <summary>Renesas Electronics opened a Physical AI and Robotics Lab in Beijing on August 27, 2026. The key signal is the company’s bill-of-materials framing: Renesas says it can address roughly 30 percent of a humanoid robot BOM today and sees an opportunity to expand coverage to 70 percent through control, power, sensing, AI, software, and ecosystem solutions. The move positions Renesas as a broad supplier platform for humanoid developers trying to move from prototypes to reliable, scalable machines.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/renesas-physical-ai-lab-beijing-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Renesas" />
    <category term="Physical AI" />
    <category term="Humanoid Robots" />
    <category term="Robotics Semiconductors" />
    <category term="Beijing" />
    <category term="Robot Supply Chain" />
    <category term="Motor Control" />
    <category term="Edge AI" />
    <category term="Power Management" />
  </entry>
  <entry>
    <title>Hyundai Wants Car Dealers to Become Robot Dealers</title>
    <link href="https://biped.news/article/hyundai-dealer-robot-channel-2026" />
    <id>biped-article-hyundai-dealer-robot-channel-2026</id>
    <updated>2026-08-27T02:05:27.808Z</updated>
    <published>2026-08-27T02:00:00.000Z</published>
    <summary>Hyundai Motor Company told investors that Boston Dynamics is moving toward Atlas mass production and that Hyundai could use dealer partners and Hyundai Capital to sell and finance robots. The idea is still exploratory, but it puts a serious industrial company behind a question the humanoid sector often avoids: commercialization will need channels, service, credit, parts, training, and customer support, not just viral demos.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/hyundai-dealer-robot-channel-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Hyundai" />
    <category term="Boston Dynamics" />
    <category term="Atlas" />
    <category term="Spot" />
    <category term="Stretch" />
    <category term="robotics distribution" />
    <category term="robot financing" />
    <category term="physical AI" />
  </entry>
  <entry>
    <title>Figure Index Turns Household Chores Into Training Data for Humanoid Robots</title>
    <link href="https://biped.news/article/figure-index-robot-training-dataset-2026" />
    <id>biped-article-figure-index-robot-training-dataset-2026</id>
    <updated>2026-09-15T18:08:43.802Z</updated>
    <published>2026-08-26T02:00:00.000Z</published>
    <summary>Figure AI launched Index, a global human-video collection network for training Helix on physical tasks. Figure reports 264,000 app downloads, more than 44,000 weekly active users, more than 16 million video uploads and $15 million paid to contributors, while its pipeline filters, fraud-checks, deduplicates, rebalances and annotates submissions. Biped.News applied through the public app to contribute household-chore data and is currently on the waiting list; this updated analysis separates those company-reported scale metrics from the still-undisclosed retained training hours and measured impact on robot performance.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://biped.news//article-images/figure-index-hero.webp" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Figure AI" />
    <category term="Figure Index" />
    <category term="Helix" />
    <category term="robot training data" />
    <category term="physical AI" />
    <category term="home robots" />
    <category term="humanoid robots" />
    <category term="household chores" />
    <category term="embodied AI" />
  </entry>
  <entry>
    <title>PaXini Tech: The Tactile Humanoid Company Behind China’s Dexterous Robot Push</title>
    <link href="https://biped.news/article/paxini-tech-company-profile" />
    <id>biped-article-paxini-tech-company-profile</id>
    <updated>2026-08-29T00:00:00.000Z</updated>
    <published>2026-08-26T01:00:00.000Z</published>
    <summary>The Shenzhen company is pairing tactile sensors, dexterous hands, and humanoid robots with fresh funding and reported IPO interest.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/paxini-tech-company-profile-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="PaXini Tech" />
    <category term="company profile" />
  </entry>
  <entry>
    <title>Humanoid Robots Beat Bolt, Then Hit the Wall</title>
    <link href="https://biped.news/article/beijing-robot-games-records-2026" />
    <id>biped-article-beijing-robot-games-records-2026</id>
    <updated>2026-08-29T00:00:00.000Z</updated>
    <published>2026-08-25T02:00:00.000Z</published>
    <summary>Humanoid robots at Beijing’s 2026 World Humanoid Robot Games reportedly ran 100 meters in 9.39 seconds and 400 meters in 38.15 seconds. The article explains why the records matter as public benchmarks for locomotion, reinforcement-learning gaits, thermal limits, and autonomy disclosure, while cautioning that sprint performance is not the same as commercial deployment readiness.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/beijing-robot-games-records-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="World Humanoid Robot Games" />
    <category term="Beijing" />
    <category term="humanoid robots" />
    <category term="physical AI" />
    <category term="robot benchmarks" />
    <category term="Tiangong Ultra" />
    <category term="Honor Lightning" />
    <category term="reinforcement learning" />
    <category term="robot autonomy" />
    <category term="robot locomotion" />
  </entry>
  <entry>
    <title>Vision-Language-Action Models Explained: How Robots Turn Camera Pixels and Commands Into Motion</title>
    <link href="https://biped.news/article/vision-language-action-models-robotics-explained" />
    <id>biped-article-vision-language-action-models-robotics-explained</id>
    <updated>2026-08-25T01:10:19.446Z</updated>
    <published>2026-08-25T01:00:00.000Z</published>
    <summary>Vision-language-action models let robots map camera images and human instructions directly into actions. The hard part is making that generalization reliable outside the demo bench.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/vision-language-action-models-robotics-explained-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="vision-language-action models" />
    <category term="physical AI" />
    <category term="RT-2" />
    <category term="OpenVLA" />
    <category term="robot learning" />
    <category term="robot manipulation" />
    <category term="foundation models" />
  </entry>
  <entry>
    <title>X Square Robot Turns WRC Into an Embodied AI Form-Factor Test</title>
    <link href="https://biped.news/article/x-square-robot-wrc-embodied-ai-2026" />
    <id>biped-article-x-square-robot-wrc-embodied-ai-2026</id>
    <updated>2026-08-29T00:00:00.000Z</updated>
    <published>2026-08-24T02:00:00.000Z</published>
    <summary>X Square Robot’s WRC 2026 showcase ties embodied AI models, task-specific arms, dexterous manipulation, home demos, and data-capture infrastructure into one strategy. The key signal is a livestreamed logistics benchmark: 1,816 parcels processed in one hour with reported accuracy above 98 percent. For humanoid robotics, the story is a form-factor reality check, useful physical AI may win first where the body is shaped around the work.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/x-square-robot-wrc-embodied-ai-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="X Square Robot" />
    <category term="WRC 2026" />
    <category term="embodied AI" />
    <category term="physical AI" />
    <category term="warehouse automation" />
    <category term="robot manipulation" />
    <category term="humanoid robots" />
  </entry>
  <entry>
    <title>Physical AI&apos;s Next Bottleneck Is the Human Operations Team</title>
    <link href="https://biped.news/article/physical-ai-workforce-bottleneck-2026" />
    <id>biped-article-physical-ai-workforce-bottleneck-2026</id>
    <updated>2026-08-23T19:19:46.080Z</updated>
    <published>2026-08-23T02:00:00.000Z</published>
    <summary>A Robot Report piece published August 22 argues that physical AI deployments are exposing a labor and operations bottleneck. Robots still need fleet operators, technicians, teleoperators, validators, and data specialists to keep real facilities moving. For humanoid robotics, the takeaway is blunt: autonomy is only one part of deployment readiness.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/physical-ai-workforce-bottleneck-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="physical AI" />
    <category term="robot operations" />
    <category term="humanoid robots" />
    <category term="teleoperation" />
    <category term="warehouse robotics" />
  </entry>
  <entry>
    <title>Schaeffler Turns Humanoid Gearboxes Into a 2027 Mass-Manufacturing Test</title>
    <link href="https://biped.news/article/schaeffler-formed-strain-wave-gearboxes-2027" />
    <id>biped-article-schaeffler-formed-strain-wave-gearboxes-2027</id>
    <updated>2026-08-22T02:00:00.000Z</updated>
    <published>2026-08-22T02:00:00.000Z</published>
    <summary>Schaeffler says it has validated formed strain-wave gearboxes for humanoid robot joints and plans to begin mass manufacturing in Germany in 2027. The company claims its forming process can reduce manufacturing cost by more than 25 percent and material use by more than 75 percent compared with conventional machining, attacking a key actuator cost center as humanoid robotics moves toward larger fleets.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/schaeffler-formed-strain-wave-gearboxes-2027-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Humanoid Robot Components" />
    <category term="Actuators" />
    <category term="Manufacturing" />
    <category term="Supply Chain" />
    <category term="Industrial Robotics" />
    <category term="Physical AI" />
  </entry>
  <entry>
    <title>Humanoid Robot Shipments Are Finally Big Enough to Need a Reality Check</title>
    <link href="https://biped.news/article/humanoid-shipment-data-reality-check-2026" />
    <id>biped-article-humanoid-shipment-data-reality-check-2026</id>
    <updated>2026-08-21T02:05:17.033Z</updated>
    <published>2026-08-21T02:00:00.000Z</published>
    <summary>Counterpoint Research says global humanoid robot shipments exceeded 22,000 units in H1 2026, up nearly 300 percent year over year, with AGIBOT and Unitree leading the early volume race. Separate China-focused reporting points to more than 40,000 Chinese shipments, highlighting a definition problem. The article explains why shipment volume matters, why it is not the same as useful work, and what buyers should demand next.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/humanoid-shipment-data-reality-check-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Counterpoint Research" />
    <category term="AGIBOT" />
    <category term="Unitree" />
    <category term="humanoid robots" />
    <category term="robot shipments" />
    <category term="physical AI" />
    <category term="China robotics" />
    <category term="deployment metrics" />
  </entry>
  <entry>
    <title>Orbbec Turns Robot Training Data Into the Main WRC Story</title>
    <link href="https://biped.news/article/orbbec-wrc-robot-free-data-physical-ai-2026" />
    <id>biped-article-orbbec-wrc-robot-free-data-physical-ai-2026</id>
    <updated>2026-08-20T02:04:45.814Z</updated>
    <published>2026-08-20T02:00:00.000Z</published>
    <summary>Orbbec’s WRC 2026 launch is not a humanoid robot reveal. It is a signal that physical AI infrastructure, especially synchronized real-world data capture, is becoming a core competitive layer for robot makers.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/orbbec-wrc-robot-free-data-physical-ai-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Orbbec" />
    <category term="World Robot Conference" />
    <category term="physical AI" />
    <category term="robotics vision" />
    <category term="robot data" />
    <category term="humanoid robots" />
    <category term="embodied AI" />
  </entry>
  <entry>
    <title>Seeing Machines Brings Driver-Monitoring DNA to Physical AI</title>
    <link href="https://biped.news/article/seeing-machines-physical-ai-platform-2026" />
    <id>biped-article-seeing-machines-physical-ai-platform-2026</id>
    <updated>2026-08-19T02:05:01.431Z</updated>
    <published>2026-08-19T02:00:00.000Z</published>
    <summary>Seeing Machines launched a Physical AI Platform on August 17, 2026, applying its human-centred AI approach from automotive driver and occupant monitoring to humanoid robotics and industrial automation. The platform builds a dynamic 3D map of people, objects, and environments so robots can interpret behavior, anticipate risk, and make safer real-time decisions around humans.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/seeing-machines-physical-ai-platform-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Seeing Machines" />
    <category term="Physical AI" />
    <category term="Humanoid Robots" />
    <category term="Human Sensing" />
    <category term="Robot Safety" />
    <category term="Computer Vision" />
    <category term="Industrial Automation" />
    <category term="Driver Monitoring" />
    <category term="Edge AI" />
  </entry>
  <entry>
    <title>Clone Robotics: The Biomimetic Android Bet</title>
    <link href="https://biped.news/article/clone-robotics-company-profile" />
    <id>biped-article-clone-robotics-company-profile</id>
    <updated>2026-08-19T01:15:38.065Z</updated>
    <published>2026-08-19T01:00:00.000Z</published>
    <summary>Clone Robotics is one of the most original humanoid startups because it is pursuing biomimetic android bodies rather than ordinary electric-actuator platforms. The company’s promise is softer, more human-compatible hardware. Its challenge is control, durability, serviceability, funding, and finding a first market where anatomical realism is worth the complexity.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/clone-robotics-company-profile-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Clone Robotics" />
    <category term="humanoid robots" />
    <category term="androids" />
    <category term="biomimetic robots" />
    <category term="synthetic muscles" />
    <category term="Protoclone" />
    <category term="Clone Alpha" />
    <category term="robotics startups" />
  </entry>
  <entry>
    <title>Foundation Pitches Phantom Humanoids for Border Security, But DHS Has No Deal Yet</title>
    <link href="https://biped.news/article/foundation-phantom-border-security-pitch-2026" />
    <id>biped-article-foundation-phantom-border-security-pitch-2026</id>
    <updated>2026-08-18T02:05:04.921Z</updated>
    <published>2026-08-18T02:00:00.000Z</published>
    <summary>Foundation is pitching its Phantom humanoid robot for U.S. border security. The company describes a reconnaissance role built around rough-terrain navigation, human detection, mapping, and alerts to officers. DHS says it has no formal deal, making this a procurement test rather than a deployment story.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/foundation-phantom-border-security-pitch-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Foundation" />
    <category term="Phantom" />
    <category term="humanoid robots" />
    <category term="border security" />
    <category term="defense robotics" />
    <category term="physical AI" />
    <category term="DHS" />
    <category term="robot surveillance" />
    <category term="autonomous systems" />
    <category term="robot procurement" />
  </entry>
  <entry>
    <title>Humanoid Robot Uptime Explained: Maintenance, Repairs, and the Real Economics of Deployment</title>
    <link href="https://biped.news/article/humanoid-uptime-maintenance-economics-explained" />
    <id>biped-article-humanoid-uptime-maintenance-economics-explained</id>
    <updated>2026-08-18T01:00:00.000Z</updated>
    <published>2026-08-18T01:00:00.000Z</published>
    <summary>Why availability, assist-event rates, modular repairs, and service contracts will decide which humanoid fleets leave pilot programs.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/humanoid-uptime-maintenance-economics-explained-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="humanoid robots" />
    <category term="robot maintenance" />
    <category term="robot uptime" />
    <category term="physical AI" />
    <category term="robot deployment" />
    <category term="RaaS" />
  </entry>
  <entry>
    <title>Beijing Turns Humanoid Robot Games Into a Physical AI Benchmark</title>
    <link href="https://biped.news/article/beijing-world-humanoid-games-2026" />
    <id>biped-article-beijing-world-humanoid-games-2026</id>
    <updated>2026-08-17T07:19:30.755Z</updated>
    <published>2026-08-17T02:00:00.000Z</published>
    <summary>Beijing is preparing to host the second World Humanoid Robot Games from August 22 to 26, 2026, with 2,056 robots from 666 teams across 16 countries and regions. The event adds tougher autonomy requirements and more practical scenarios across factories, hotels, homes, logistics, and emergency response. The article explains why the games matter as a public benchmark, why China has a scale advantage in competition data, and why competition success should not be confused with commercial deployment proof.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/beijing-world-humanoid-games-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="World Humanoid Robot Games" />
    <category term="Beijing" />
    <category term="humanoid robots" />
    <category term="physical AI" />
    <category term="robot benchmarks" />
    <category term="China robotics" />
    <category term="robot autonomy" />
    <category term="robot competitions" />
    <category term="embodied AI" />
    <category term="robot standards" />
  </entry>
  <entry>
    <title>Mitsubishi and Highlanders Put Humanoid Production on the Old Engine Plant Clock</title>
    <link href="https://biped.news/article/mitsubishi-highlanders-kyoto-humanoid-production-2026" />
    <id>biped-article-mitsubishi-highlanders-kyoto-humanoid-production-2026</id>
    <updated>2026-08-16T02:05:33.107Z</updated>
    <published>2026-08-16T02:00:00.000Z</published>
    <summary>Mitsubishi Motors and Highlanders have signed an MOU to jointly develop humanoid robots for Mitsubishi factories and explore production at the automaker&apos;s Kyoto Plant. The plan is not a confirmed production launch yet, but it gives the humanoid sector something more concrete than a demo: an automaker with idle plant space, quality systems, safety engineering, and a reason to use the robots internally before selling them broadly.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/mitsubishi-highlanders-kyoto-humanoid-production-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Mitsubishi Motors" />
    <category term="Highlanders" />
    <category term="humanoid robots" />
    <category term="physical AI" />
    <category term="Japan" />
    <category term="manufacturing" />
  </entry>
  <entry>
    <title>LG and NVIDIA Put Humanoid Robots on the Electronics Giant Track</title>
    <link href="https://biped.news/article/lg-nvidia-humanoid-robot-stack-2026" />
    <id>biped-article-lg-nvidia-humanoid-robot-stack-2026</id>
    <updated>2026-08-15T02:06:36.975Z</updated>
    <published>2026-08-15T02:00:00.000Z</published>
    <summary>LG Electronics announced plans to unveil a next-generation bipedal humanoid robot in Q1 2027 through an expanded NVIDIA collaboration. The planned robot will use NVIDIA Isaac GR00T and Jetson Thor while drawing on LG hardware capabilities across sensors, batteries, actuators, and manufacturing. The story is less about another humanoid teaser and more about whether large electronics groups can turn physical AI stacks into deployable products.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/lg-nvidia-humanoid-robot-stack-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="LG Electronics" />
    <category term="NVIDIA" />
    <category term="Jetson Thor" />
    <category term="Isaac GR00T" />
    <category term="Humanoid Robots" />
    <category term="Physical AI" />
    <category term="Robotics Compute" />
    <category term="Robot Manufacturing" />
  </entry>
  <entry>
    <title>13 Humanoid Deployments, Zero Complete Public Safety Files</title>
    <link href="https://biped.news/data-briefs/humanoid-deployment-safety-transparency-gap" />
    <id>biped-data-brief-2026-08-safety-transparency-gap</id>
    <updated>2026-08-14T12:00:00.000Z</updated>
    <published>2026-08-14T12:00:00.000Z</published>
    <summary>Biped.News reviewed 13 current core humanoid deployment/test records and found zero that publicly disclose all five buyer-critical safety artifacts in its August 14, 2026 transparency checklist.</summary>
    <link rel="enclosure" type="image/png" href="https://biped.news/data-briefs/humanoid-deployment-safety-transparency-gap-social.png" />
    <author><name>Biped News</name></author>
    <category term="Biped Data Brief" />
    <category term="Original Research" />
  </entry>
  <entry>
    <title>Honda Puts ASIMO DNA Back to Work With Avatar Robot Hand</title>
    <link href="https://biped.news/article/honda-avatar-hand-teleoperation-2026" />
    <id>biped-article-honda-avatar-hand-teleoperation-2026</id>
    <updated>2026-08-14T02:05:30.052Z</updated>
    <published>2026-08-14T02:00:00.000Z</published>
    <summary>Honda’s latest Avatar Robot update is not a mass deployment claim. It is a technical signal that teleoperation and dexterous manipulation may be a faster path to useful humanoid work than waiting for full autonomy.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/honda-avatar-hand-teleoperation-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Honda" />
    <category term="Avatar Robot" />
    <category term="ASIMO" />
    <category term="humanoid robots" />
    <category term="robotic hands" />
    <category term="teleoperation" />
    <category term="physical AI" />
  </entry>
  <entry>
    <title>San Mateo County Puts Humanoid Robots on the Permit Clock</title>
    <link href="https://biped.news/article/san-mateo-humanoid-robot-rules-2026" />
    <id>biped-article-san-mateo-humanoid-robot-rules-2026</id>
    <updated>2026-08-13T02:04:48.024Z</updated>
    <published>2026-08-13T02:00:00.000Z</published>
    <summary>San Mateo County supervisors approved an early local regulatory framework for commercial humanoid robot use. The resolution points toward permits, emergency kill switches, fire safety planning, and tracking of possible job displacement. The article explains why this matters for robotics companies, buyers, local governments, and the safety stack that will shape humanoid deployment.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/san-mateo-humanoid-robot-rules-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="San Mateo County" />
    <category term="humanoid robots" />
    <category term="robot policy" />
    <category term="robot safety" />
    <category term="physical AI" />
    <category term="commercial robots" />
    <category term="lithium-ion batteries" />
    <category term="robot permits" />
  </entry>
  <entry>
    <title>AgiBot Passes Unitree as China&apos;s Humanoid Shipment Race Gets Real</title>
    <link href="https://biped.news/article/agibot-unitree-h1-shipments-china-2026" />
    <id>biped-article-agibot-unitree-h1-shipments-china-2026</id>
    <updated>2026-08-12T02:06:14.406Z</updated>
    <published>2026-08-12T02:00:00.000Z</published>
    <summary>Smart Analytics Global says AgiBot shipped about 8,400 humanoid robots in the first half of 2026, overtaking Unitree at about 5,900 units. The report puts global H1 shipments near 19,100 units, up 272 percent year over year, with Chinese vendors accounting for more than 97 percent of shipments. The article explains why the data matters, why shipments are not the same as useful work, and what buyers should demand next.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/agibot-unitree-h1-shipments-china-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="AgiBot" />
    <category term="Unitree" />
    <category term="China" />
    <category term="humanoid robots" />
    <category term="robot shipments" />
    <category term="physical AI" />
    <category term="embodied AI" />
    <category term="robotics IPO" />
  </entry>
  <entry>
    <title>Pudu Robotics: The Service Robot Company Walking Into Humanoids</title>
    <link href="https://biped.news/article/pudu-robotics-company-profile" />
    <id>biped-article-pudu-robotics-company-profile</id>
    <updated>2026-08-12T12:12:03.087Z</updated>
    <published>2026-08-12T01:00:00.000Z</published>
    <summary>Company profile of Pudu Robotics, covering its 2016 founding, service robot fleet, D9 humanoid, 2026 financing, global channels, and commercial humanoid strategy.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/pudu-robotics-company-profile-hero.png" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Pudu Robotics" />
    <category term="PUDU D9" />
    <category term="humanoid robots" />
    <category term="service robots" />
    <category term="embodied AI" />
    <category term="commercial robotics" />
  </entry>
  <entry>
    <title>Electronic Skin Gives Humanoid Robots a Touch-Sensing Path</title>
    <link href="https://biped.news/article/hanyang-electronic-skin-robot-touch-sensing-2026" />
    <id>biped-article-hanyang-electronic-skin-robot-touch-sensing-2026</id>
    <updated>2026-08-11T02:05:09.015Z</updated>
    <published>2026-08-11T02:05:09.015Z</published>
    <summary>Hanyang University ERICA researchers have reported a vertically integrated dual-gate tribotronic transistor for electronic skin. The architecture uses a PDMS triboelectric sensing layer, gate insulator, and ITZO thin-film transistor to support tunable sensitivity, contact detection, pressure response, proximity sensing, and dense array integration. For humanoid robotics, the signal is tactile infrastructure: robot hands and body surfaces need contact data that cameras cannot reliably infer.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/hanyang-electronic-skin-robot-touch-sensing-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="electronic skin" />
    <category term="tactile sensing" />
    <category term="robot sensors" />
    <category term="humanoid robots" />
    <category term="physical AI" />
    <category term="Hanyang University ERICA" />
    <category term="Jaekyun Kim" />
    <category term="tribotronic transistor" />
    <category term="prosthetics" />
    <category term="robot hands" />
    <category term="sensor arrays" />
  </entry>
  <entry>
    <title>Impedance Control Explained: How Humanoid Robots Learn Safe Contact</title>
    <link href="https://biped.news/article/impedance-control-humanoid-robots-explained" />
    <id>biped-article-impedance-control-humanoid-robots-explained</id>
    <updated>2026-08-12T16:14:37.690Z</updated>
    <published>2026-08-11T01:00:00.000Z</published>
    <summary>Impedance control shapes the relationship between force and motion so humanoid robots can manage contact, balance, bracing, and safe interaction instead of only moving through empty space.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/impedance-control-humanoid-robots-explained-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="impedance control" />
    <category term="humanoid robots" />
    <category term="compliance" />
    <category term="force control" />
    <category term="physical AI" />
    <category term="whole-body control" />
  </entry>
  <entry>
    <title>Laundry Is Becoming the Home-Robot Proving Ground</title>
    <link href="https://biped.news/article/laundry-home-robot-proving-ground-2026" />
    <id>biped-article-laundry-home-robot-proving-ground-2026</id>
    <updated>2026-08-10T21:09:18.275Z</updated>
    <published>2026-08-10T12:10:26.629Z</published>
    <summary>Biped compares Sunday Memo, Weave Isaac 1, and 1X NEO through the emerging laundry benchmark. The feature examines Sunday’s company-reported 99.1% ACT-2 evaluation across 785 autonomous attempts, Weave’s $7,999/$449-per-month Laundry Flow product, 1X’s $20,000/$499-per-month NEO and Expert Mode, the technical challenge of deformable-object manipulation, and a proposed Biped basket-to-drawer benchmark for full-load completion, intervention minutes, speed, quality, privacy, and cost per useful hour.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/laundry-home-robot-proving-ground-2026-hero.png" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="laundry robot" />
    <category term="home robots" />
    <category term="humanoid robots" />
    <category term="Sunday Robotics" />
    <category term="Memo" />
    <category term="Weave Robotics" />
    <category term="Isaac 1" />
    <category term="1X" />
    <category term="NEO" />
    <category term="physical AI" />
    <category term="deformable object manipulation" />
    <category term="household robots" />
    <category term="robot reliability" />
    <category term="teleoperation" />
  </entry>
  <entry>
    <title>Shenzhen Shows Humanoid Robots at Work, But the Data Gap Is Still the Story</title>
    <link href="https://biped.news/article/shenzhen-humanoid-service-rollout-2026" />
    <id>biped-article-shenzhen-humanoid-service-rollout-2026</id>
    <updated>2026-08-10T03:18:34.737Z</updated>
    <published>2026-08-10T02:00:00.000Z</published>
    <summary>Xinhua published an August 9 report showing humanoid robots in several Shenzhen settings, including restaurant delivery, China Post package sorting, Futian traffic guidance, and ice cream service at a park. The article analyzes why the report matters for physical AI, what is confirmed, what remains unverified, and why Shenzhen&apos;s public-facing rollout should be judged by intervention rate, uptime, safety evidence, and repeatable economics rather than by photos alone.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/shenzhen-humanoid-service-rollout-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Shenzhen" />
    <category term="China" />
    <category term="humanoid robots" />
    <category term="physical AI" />
    <category term="logistics" />
    <category term="service robots" />
    <category term="deployment" />
  </entry>
  <entry>
    <title>Real World AI Gets Its Own Stage as Humanoid Robotics Hits the Production Test</title>
    <link href="https://biped.news/article/real-world-ai-stage-production-gap-2026" />
    <id>biped-article-real-world-ai-stage-production-gap-2026</id>
    <updated>2026-08-09T02:00:00.000Z</updated>
    <published>2026-08-09T02:00:00.000Z</published>
    <summary>TechCrunch has created a Real World AI Stage for Disrupt 2026, with sessions focused on safety-critical systems, edge AI, and the gap between prototype, product, and scaled business. The article explains why that programming shift matters for humanoid robotics: the category is entering a production-readiness phase where validation, local compute, support burden, deployment metrics, and customer economics matter more than isolated demo videos.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/real-world-ai-stage-production-gap-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="physical AI" />
    <category term="humanoid robots" />
    <category term="TechCrunch Disrupt" />
    <category term="Real World AI" />
    <category term="edge AI" />
    <category term="robotics safety" />
    <category term="industrial automation" />
    <category term="robot deployment" />
    <category term="robotics startups" />
    <category term="field operations" />
  </entry>
  <entry>
    <title>Unitree Prices Shanghai IPO at $9 Billion Valuation as DeepSeek Invests</title>
    <link href="https://biped.news/article/unitree-ipo-priced-9b-deepseek-2026" />
    <id>biped-article-unitree-ipo-priced-9b-deepseek-2026</id>
    <updated>2026-08-08T12:30:00.000Z</updated>
    <published>2026-08-08T12:30:00.000Z</published>
    <summary>Unitree has moved from filing to final IPO terms: RMB 150.80 per share, about RMB 61 billion ($9.04 billion) implied valuation, roughly RMB 6.1 billion in expected proceeds, and a RMB 140.8 million strategic investment from DeepSeek. The shares had not yet been verified as trading as of August 8, so Biped.News treats this as a priced offering rather than a completed market debut.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/unitree-ipo-shanghai-610m-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Unitree Robotics" />
    <category term="IPO" />
    <category term="STAR Market" />
    <category term="Shanghai Stock Exchange" />
    <category term="DeepSeek" />
    <category term="China" />
    <category term="humanoid robots" />
    <category term="robotics investing" />
  </entry>
  <entry>
    <title>12 of 13 Tracked Humanoid Deployments Are Still Pilots or Tests</title>
    <link href="https://biped.news/data-briefs/humanoid-deployment-pilot-production-gap" />
    <id>biped-data-brief-2026-08-deployment-maturity-gap</id>
    <updated>2026-08-08T12:00:00.000Z</updated>
    <published>2026-08-08T12:00:00.000Z</published>
    <summary>Biped.News analyzed 13 current core humanoid deployment/test records: nine pilots, three testing records, and one production-status commercial operation.</summary>
    <link rel="enclosure" type="image/png" href="https://biped.news/data-briefs/humanoid-deployment-pilot-production-gap-social.png" />
    <author><name>Biped News</name></author>
    <category term="Biped Data Brief" />
    <category term="Original Research" />
  </entry>
  <entry>
    <title>Icarus Robotics Wants to Train Physical AI in Orbit</title>
    <link href="https://biped.news/article/icarus-space-robotics-iss-2026" />
    <id>biped-article-icarus-space-robotics-iss-2026</id>
    <updated>2026-08-07T02:04:39.499Z</updated>
    <published>2026-08-07T02:00:00.000Z</published>
    <summary>A new Robotics &amp; Automation News interview with Icarus Robotics CTO Jamie Palmer puts space robotics into the physical AI conversation. The company, founded in 2024, is building robotic workers for cargo logistics, inspection, and experiment support in orbit. Its plan starts with teleoperation, uses ISS work to collect real microgravity data, and moves toward shared autonomy as commercial space stations, orbital manufacturing, and lunar infrastructure create more work than astronauts can absorb.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/icarus-space-robotics-iss-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Icarus Robotics" />
    <category term="space robotics" />
    <category term="physical AI" />
    <category term="embodied AI" />
    <category term="ISS" />
    <category term="teleoperation" />
    <category term="microgravity" />
  </entry>
  <entry>
    <title>AMD Enters the Robot Brain Race With Kria AI and Ryzen Embedded X100</title>
    <link href="https://biped.news/article/amd-kria-robotics-compute-2026" />
    <id>biped-article-amd-kria-robotics-compute-2026</id>
    <updated>2026-08-06T02:00:00.000Z</updated>
    <published>2026-08-06T02:00:00.000Z</published>
    <summary>AMD has launched Ryzen AI Embedded X100 processors, Kria AI system-on-modules, and a Kria AI Robotics Developer Platform for physical AI systems. The article explains why consolidated edge compute matters for humanoid robots, how AMD is positioning against NVIDIA, and what evidence robot makers and buyers should demand before treating the platform as production-ready.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/amd-kria-robotics-compute-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="AMD" />
    <category term="Kria AI" />
    <category term="Ryzen AI Embedded X100" />
    <category term="Physical AI" />
    <category term="Robotics Compute" />
    <category term="Humanoid Robots" />
  </entry>
  <entry>
    <title>Agile Robots Turns Factory Revenue Into a Physical AI Signal</title>
    <link href="https://biped.news/article/agile-robots-revenue-industrial-ai-2026" />
    <id>biped-article-agile-robots-revenue-industrial-ai-2026</id>
    <updated>2026-08-05T02:00:00.000Z</updated>
    <published>2026-08-05T02:00:00.000Z</published>
    <summary>Agile Robots is reported to expect 2026 revenue to roughly double from about EUR 300 million in 2025, with profitability targeted within two to three years. The article explains why that commercial base matters for physical AI, how the Google DeepMind partnership fits, and why Agile ONE should be read as part of a broader factory automation strategy rather than a standalone humanoid deployment story.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/agile-robots-revenue-industrial-ai-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Agile Robots" />
    <category term="Google DeepMind" />
    <category term="Gemini Robotics" />
    <category term="physical AI" />
    <category term="industrial automation" />
    <category term="humanoid robots" />
    <category term="Agile ONE" />
    <category term="factory robotics" />
    <category term="thyssenkrupp Automation Engineering" />
  </entry>
  <entry>
    <title>Mujin: The Robot Intelligence Layer Behind Industrial Automation</title>
    <link href="https://biped.news/article/mujin-company-profile" />
    <id>biped-article-mujin-company-profile</id>
    <updated>2026-08-05T01:00:00.000Z</updated>
    <published>2026-08-05T01:00:00.000Z</published>
    <summary>Company profile of Mujin, covering its 2011 founding, MujinOS platform, $233 million Series D, reported $411 million total funding, warehouse and factory automation use cases, and relevance to humanoid and embodied AI systems.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/mujin-company-profile-hero.png" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Mujin" />
    <category term="MujinOS" />
    <category term="industrial robots" />
    <category term="motion planning" />
    <category term="warehouse automation" />
    <category term="physical AI" />
  </entry>
  <entry>
    <title>BYD&apos;s First Humanoid Robot Is Starting in the Showroom</title>
    <link href="https://biped.news/article/byd-di-space-humanoid-showroom-2026" />
    <id>biped-article-byd-di-space-humanoid-showroom-2026</id>
    <updated>2026-08-04T18:29:34.450Z</updated>
    <published>2026-08-04T02:00:00.000Z</published>
    <summary>BYD plans to unveil its first humanoid robot in early August at Zhengzhou Di Space. The article explains why a showroom-first robot is a serious test of customer interaction, supervised autonomy, and automaker robotics strategy, while cautioning that this remains an announced prototype rather than a proven commercial rollout.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/byd-di-space-humanoid-showroom-2026-1.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="BYD" />
    <category term="Di Space" />
    <category term="Xiao Di" />
    <category term="humanoid robots" />
    <category term="EV robotics" />
    <category term="physical AI" />
    <category term="showroom automation" />
    <category term="China robotics" />
    <category term="Tesla Optimus" />
  </entry>
  <entry>
    <title>Humanoid Actuators Explained: The Motor, Gearbox, and Torque Stack Behind Physical AI</title>
    <link href="https://biped.news/article/humanoid-actuators-explained" />
    <id>biped-article-humanoid-actuators-explained</id>
    <updated>2026-08-12T16:14:37.321Z</updated>
    <published>2026-08-04T01:00:00.000Z</published>
    <summary>Sim-to-real transfer lets humanoid robots practice risky movements in physics engines, then adapt those policies to contact, friction, latency, and hardware limits.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/humanoid-actuators-explained-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Sim-to-Real" />
    <category term="Physical AI" />
    <category term="Humanoid Robots" />
    <category term="Robot Training" />
    <category term="Control" />
  </entry>
  <entry>
    <title>China Pushes Back on the U.S. Humanoid Robot Ban</title>
    <link href="https://biped.news/article/china-responds-us-humanoid-robot-ban-2026" />
    <id>biped-article-china-responds-us-humanoid-robot-ban-2026</id>
    <updated>2026-08-03T10:51:23.111Z</updated>
    <published>2026-08-03T14:00:00.000Z</published>
    <summary>China&apos;s response to the FCC&apos;s July 28 advanced-robotics action turns the U.S. humanoid robot ban into a wider trade and industrial-policy fight. Beijing argues the rule overextends national-security claims, while Chinese state-linked commentary says the ban could raise U.S. costs, slow access to low-cost robot hardware, and push Chinese robotics suppliers toward other markets.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/fcc-covered-list-foreign-robots-2026-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="China" />
    <category term="FCC" />
    <category term="humanoid robots" />
    <category term="robotics policy" />
    <category term="Unitree" />
    <category term="physical AI" />
    <category term="supply chain security" />
    <category term="trade policy" />
  </entry>
  <entry>
    <title>Tau Robotics Turns Home Cleaning Into a Humanoid Data Business</title>
    <link href="https://biped.news/article/tau-robotics-humanoid-cleaning-service-2026" />
    <id>biped-article-tau-robotics-humanoid-cleaning-service-2026</id>
    <updated>2026-08-03T11:15:03.019Z</updated>
    <published>2026-08-03T02:00:00.000Z</published>
    <summary>Tau Robotics has opened an invite-only humanoid cleaning service in San Francisco at $30 for a one-hour visit. The company is not selling robots, it is selling a service, with AI and remote human operators working together while the robot performs limited cleaning tasks. This article explains why the service model matters, what is confirmed, what remains unverified, and how it fits into the larger shift from humanoid demos to supervised real-world data collection.</summary>
    <link rel="enclosure" type="image/jpeg" href="https://storage.googleapis.com/biped-f5a14-media/article-images/tau-hero.jpg" />
    <author>
      <name>Biped News</name>
    </author>
    <category term="Tau Robotics" />
    <category term="humanoid cleaning" />
    <category term="home robots" />
    <category term="physical AI" />
    <category term="teleoperation" />
    <category term="robotics as a service" />
    <category term="San Francisco robotics" />
    <category term="embodied AI" />
    <category term="robot deployment" />
  </entry>
</feed>