Tag: Humanoid Robots

  • Humanoid Robots Compete in Beijing Half-Marathon: Batteries vs. Muscles

    Humanoid Robots Compete in Beijing Half-Marathon: Batteries vs. Muscles

    Key Takeaways

    1. Humanoid robots are expected to increasingly assist in tasks across healthcare, logistics, and construction.
    2. A key requirement for these robots is the ability to walk steadily on two legs over long distances in real environments.
    3. The robots participated in a half marathon in Beijing to test their technical endurance, focusing on durability, energy efficiency, and battery performance.
    4. The race covered about 13 miles, with the robots prioritizing stability and endurance rather than speed.
    5. The fastest robot, “Tiangong,” completed the race in 2:40:24, showing potential but still lacking the efficiency for everyday tasks.


    In the coming years, it’s anticipated that humanoid robots will play a larger role in assisting or even taking over tasks in areas like healthcare, logistics, and construction. A vital requirement for this development is their ability to walk steadily on two legs, even over extended distances and in actual environments. This capability was put to the test in Beijing, where humanoid robots participated in a half marathon for the very first time.

    Technical Endurance Test

    Werner Kraus from Fraunhofer IPA shared with the German site ingenieur.de that the robots’ involvement was primarily a test of technical endurance. The main aim was to evaluate the durability and energy efficiency of their drives, as well as battery performance. One significant challenge is achieving a balance between runtime and weight—bigger batteries can provide greater range but also increase weight, which in turn can lower overall efficiency.

    The Marathon Course

    The race spanned approximately 13 miles and wound through the “E-Town” innovation area in southern Beijing. While human runners aimed for the best times, the robots had a different purpose: to simply complete the race. Speed wasn’t their main concern; instead, the emphasis was on stability, endurance, and reliable performance. The robots varied in height from 30 to 71 inches and weighed as much as 194 pounds, running either independently or with help from their creators.

    Race Results

    The fastest human finished the race in 1:11:07, while the quickest robot, called “Tiangong,” crossed the finish line in 2:40:24. Other robots took significantly longer. Some struggled, wobbled, or could only move forward with assistance. Nevertheless, the event demonstrated that humanoid robots can traverse long distances, even if they still lack the efficiency and reliability needed for regular, everyday tasks. A few exciting moments from the race can be watched in the livestream video linked below:

    Source:
    Link


  • China’s Smart Robotics Industry Booms with 450,000+ Companies

    China’s Smart Robotics Industry Booms with 450,000+ Companies

    Key Takeaways

    1. China’s smart robotics industry has registered 451,700 companies and reached a capital of 6.44 trillion yuan ($884.27 billion) by December 2024, with a growth of 206.73% since 2020.

    2. The eastern region of China dominates the smart robotics sector, hosting over two-thirds of companies, with a focus on scientific research, technical services, IT, and wholesale.

    3. Significant advancements in humanoid robotics are being made by companies like UBTech and Unitree Robotics, with major investments and partnerships in the manufacturing sector.

    4. The Chinese government is actively promoting robotics technology through policies aimed at fostering innovation, supporting startups, and developing supply chains, with a goal to lead the humanoid robotics industry by 2025.

    5. The humanoid robot market in China is projected to reach 300 billion yuan by 2035, solidifying the country’s position as a global leader in this area.


    China’s Rising Smart Robotics Sector

    China’s smart robotics industry is rapidly expanding. By December 2024, the country has seen the registration of 451,700 companies in this field, amounting to a total capital of 6.44 trillion yuan ($884.27 billion). Since 2020, the number of businesses has surged by 206.73%, indicating a consistent growth pattern. In just the last year, the sector has grown by 19.39%, reflecting a robust demand and investment in robotic technologies.

    Geographical Distribution and Industry Focus

    The majority of smart robotics companies in China are involved in scientific research, technical services, IT, and wholesale, collectively representing nearly 80% of the industry. The eastern part of China leads the robotics scene, hosting more than two-thirds of all enterprises in the field. This region is often referred to as the “golden cluster zone” because of its advanced research and development facilities. Additionally, the central and northwestern regions also have a notable presence, accounting for 15.33% and 14.97% of the total companies, respectively.

    Advancements in Humanoid Robotics

    China is advancing significantly in humanoid robotics. Firms such as UBTech, Unitree Robotics, CloudMinds, and Engine AI are making notable progress with AI-driven humanoid robots. UBTech has teamed up with Foxconn to integrate robots into the manufacturing process, while Unitree launched its G1 humanoid robot in July 2024. Major automakers like BYD and Geely are also adopting humanoid robots in their production lines. Investment in this sector is substantial, with China capturing 56 out of 69 global financing events in 2024, amounting to 11 billion yuan ($1.5 billion).

    Government Initiatives to Drive Progress

    The Chinese government has placed a strong emphasis on advancing robotics technology. By 2025, the goal is to create a leading humanoid robotics industry with mass production abilities. Various policies in cities like Beijing and Zhejiang are aimed at promoting innovation in robotics, assisting startups, and enhancing supply chains. By 2035, the humanoid robot market in China is predicted to reach 300 billion yuan, reinforcing its status as a worldwide leader in this domain.

  • OpenAI Expands into Humanoid Robots, AI Chips, and Smart Jewelry

    OpenAI Expands into Humanoid Robots, AI Chips, and Smart Jewelry

    OpenAI submitted a fresh trademark application to the United States Patent and Trademark Office (USPTO) on January 31, 2025. This move indicates that OpenAI is looking to grow into new areas, such as hardware and robotics. While the application mentions possible new products, it’s essential to understand that these innovations might take a few years before they are available to consumers.

    What Products Are on the Horizon?

    The trademark application lists several hardware products, including headphones, glasses, smartwatches, smart jewelry, and cases for laptops and phones. OpenAI also suggests that it may enter the virtual reality (VR) and augmented reality (AR) sectors with new headsets. These products are intended to be compatible with AI technologies, aiming to improve user experiences through AI-enhanced simulations and training.

    Collaborations and Production Timeline

    OpenAI CEO Sam Altman has stated that the company is collaborating with various firms to introduce AI-focused consumer electronics to the market. Nevertheless, it is still uncertain when these products will be ready, as prototypes are anticipated to require several years for development.

    Advancements in Robotics

    In addition to hardware, the trademark application indicates that OpenAI aims to explore robotics, particularly humanoid robots. These robots will be programmable by users and will possess sophisticated communication and learning capabilities. To aid this vision, OpenAI has established a new robotics team led by Caitlin Kalinowski, who previously worked in Meta’s AR glasses sector. This team will concentrate on building robots equipped with tailored sensors and AI to replicate human intelligence and behavior effectively.

    Exploring AI Chips and Quantum Tech

    OpenAI is also looking into creating custom AI chips to enhance the performance of its AI models. The company is collaborating with Broadcom and TSMC, aiming to launch these chips by 2026. Furthermore, OpenAI is investigating quantum computing to utilize this technology for optimizing AI models in the future.

    While the trademark application showcases OpenAI’s aspirations in hardware, robotics, AI chips, and quantum computing, it’s still unclear when these advancements will become available to the public. The company is dedicated to developing AI-powered products, yet a significant amount of the work remains in the initial phases.

  • Figure AI’s Humanoid Robot Delivers to First Customer

    Figure AI’s Humanoid Robot Delivers to First Customer

    Figure 02 is the newer model of Figure 01, which was launched in October 2023, and it was first shown to the public in August 2024. This humanoid robot, developed by the Californian start-up Figure AI, is outfitted with six RGB cameras, depth cameras, and infrared sensors, providing it with an impressive full 360-degree view of its surroundings. Its hands have 16 degrees of freedom, allowing it to replicate human-like precision and strength, making it ideal for detailed tasks that require both hands.

    Advanced Features

    Thanks to a partnership with OpenAI, the minds behind ChatGPT, Figure 02 can comprehend voice commands and make its own choices, enabling it to work as a “thinking” companion. During a trial at the BMW factory in Spartanburg, USA, the robot showed off its practical abilities. The accompanying video gives a sneak peek into how well it performed during these tests.

    Exciting Developments

    Brett Adcock, the founder and CEO of Figure AI, announced on LinkedIn that Figure 02 is already “hard at work”, hinting at its initial delivery to a customer. However, specifics about where it will be deployed and its price haven’t been shared yet. This achievement marks a key moment in the journey toward making humanoid robots more commercially available.

    Strong Backing

    Figure AI is supported by major investors such as Microsoft, OpenAI, Nvidia, and Jeff Bezos, which has led to a valuation of $2.6 billion, placing it among the best-funded startups in the industry. The first delivery is likely to bring the company nearer to Tesla, whose Optimus robot is still limited to usage within its factories. Notably, Figure AI has transitioned from its establishment in January 2022 to monetizing its product in only 31 months.

    Source: Link

  • Nio Testing HarmonyOS Humanoid Robots: Report

    Nio Testing HarmonyOS Humanoid Robots: Report

    Chinese electric vehicle (EV) manufacturer Nio is reportedly exploring the potential of humanoid robots in its production lines. A recent report by China Star Market revealed that Nio is testing the Kuavo robot, developed by Leju Robotics, at its factories.

    Notably, Kuavo is the first HarmonyOS-powered robot unveiled by Huawei at the HDC 2024 developer conference in June.

    Nio is not the only car manufacturer exploring the use of humanoid robots in assembly lines. However, it appears to be leading this development.

    Previous Initiatives

    Prior to testing Kuavo, another humanoid robot, the Walker S from UBTech Robotics, underwent training at a Nio factory in February 2024. This marked the first instance of real-world training for the Walker S robot within a new energy vehicle (NEV) factory setting.

    During the training, the Walker S utilized its image capture and transmission capabilities for quality control tasks like inspecting door locks, seat belts, and headlight covers. Additionally, the robot was able to affix vehicle logos, as shown in a video shared by UBTech Robotics.

    Focus on Humanoid Robotics

    Nio’s interest in humanoid robotics extends beyond individual robots. In April 2024, Yi Peng, Nio’s head of humanoid robots for intellectual engineering, revealed the formation of a “squad” of humanoid robots at the China Humanoid Robot Ecology Conference. Yi also outlined two key areas of focus for Nio’s humanoid robotics endeavors:

    1. Fixing humanoid robotic arms in tight spaces and multitasking with 5G and vision algorithms.
    2. Using humanoid robots in some quality control processes.

    While the extent of Nio’s collaboration with Huawei and Leju Robotics remains unclear, the testing of the HarmonyOS-powered Kuavo robot is a significant step in the company’s efforts to incorporate humanoid robots into its manufacturing processes.

  • NVIDIA CEO Jensen Huang Foresees Popular Adoption of Humanoid Robots

    NVIDIA CEO Jensen Huang Foresees Popular Adoption of Humanoid Robots

    In a recent conversation, NVIDIA CEO Jensen Huang discussed his forward-looking perspective on the future, foreseeing a scenario where humanoid robots will soon become ubiquitous across different sectors, bringing about a significant transformation.

    Vision for Mainstream Humanoid Robots

    Huang highlighted that the production costs associated with humanoid robots are poised to be remarkably affordable, comparable to the pricing brackets of budget-friendly automobiles. He suggested a price span of $10,000 to $20,000 for these robots, aiming to make them widely accessible to a diverse consumer base.

    Potential Impact on Industries

    Drawing an analogy to the automotive realm, Huang underscored the adaptability and versatility that robots could offer within specific human-created environments. These robots are envisioned to tackle hazardous, repetitive, or monotonous tasks, ultimately boosting efficiency and safety within workplaces.

    Supporting Huang's perspective, Tesla CEO Elon Musk also mentioned that the manufacturing outlay for the humanoid robot, Optimus, is merely half of what is needed for Tesla electric vehicles. Musk envisioned an initial pricing strategy for Optimus ranging from $25,000 to $30,000, with anticipated price reductions over time.

    Technological Advancements by NVIDIA

    In line with their futuristic outlook, NVIDIA introduced Project GR00T, a foundational model tailored for humanoid robots, and launched Jetson Thor, a robust computing system specially crafted for these robots. Leveraging NVIDIA's Thor system-on-a-chip (SoC), Jetson Thor offers cutting-edge performance, energy efficiency, and a compact form factor, propelling the evolution of humanoid robotics.

    Moreover, NVIDIA's collaborations with prominent humanoid robot firms like 1X Technologies, Agility Robotics, and Boston Dynamics signify a concerted endeavor to build comprehensive AI frameworks for robotics.

    Empowering Developers and Robotics Advancements

    NVIDIA's enhancements to the Isaac platform empower developers to design and deploy sophisticated robots across various settings. Tools such as Isaac Lab for reinforcement learning and OSMO for compute orchestration facilitate the rapid enhancement of robotic capabilities. Furthermore, offerings like Isaac Manipulator and Isaac Perceptor provide pre-trained models and advanced features for robotic arms and self-governing mobile robots.

    Huang's visionary outlook, combined with NVIDIA's technological progress, paints a picture of a future where humanoid robots seamlessly integrate into our daily routines, reshaping industries and augmenting human capabilities.

    (source: NVIDIA)