A single destination for timely, editor-curated robotics news from around the world.
ACE Robotics and NTU S-Lab have introduced Mundo, a groundbreaking tool that transforms one to twelve photos of real-world locations into editable 3D scenes. This innovative technology incorporates physics to facilitate robot navigation and manipulation within these virtual environments. The signif...
Tencent Cloud has introduced a comprehensive multimodal data processing and model training stack aimed at enhancing embodied AI and autonomous driving capabilities. This initiative is significant as it addresses the growing demand for advanced AI solutions in robotics and self-driving technologies,...
NVIDIA has launched Flora, a suite of models designed to convert robot and human videos into labeled training data for robotics AI. This innovative system identifies specific actions, describes them linguistically, and evaluates task performance, facilitating the transformation of extensive video co...
Hesai has introduced Kosmo, a 3D world data platform that integrates the Kosmo One spatial camera and Kosmo World cloud service to transform real-world environments into digital formats for robot training and world modeling. An operator can scan a typical 100-square-meter indoor area in approximatel...
Mecka has successfully raised $60 million in a Series B funding round led by Sequoia Capital, with participation from Nvidia, Microsoft’s M12, Qualcomm Ventures, and Samsung. This funding, announced on October 7, aims to bolster Mecka's efforts in gathering human demonstrations for training robot mo...
X Square Robot's TwinDEX has successfully completed a 24-step chemistry experiment using a unique combination of a wearable exoskeleton and a three-finger hand. This achievement is notable as the policy governing the robot's actions was trained solely on a few hundred demonstrations that did not inv...
The rapid increase in robot model parameters has not been matched by the availability of high-quality physical interaction data. Current compliant data in China stands at only 500,000 hours, while commercial deployment requires tens of millions of hours, resulting in a gap exceeding 99%. The China A...
On September 23, Chen Pu, Vice President of Product Development at Yuan Ke Vision, highlighted the importance of high-quality real data for robot training during a seminar in Beijing. He emphasized the need to elevate data from 2D to 4D and improve precision from centimeter to sub-millimeter levels...
On September 23, Mifengpai launched a global initiative to crowdsource data for robot training, showcasing over 50,000 real environments and 5,000 tasks. The event highlighted the need for extensive data on everyday actions, which are crucial for training robots but have not been systematically reco...
Shangpin Home, in collaboration with Tangyuan Technology, has launched the WorldSimReady-Home simulation dataset aimed at addressing the challenges of robotic training in complex home environments. This open-source dataset includes 100,000 square meters of high-fidelity home scenes, 10,000 interacti...
Kinetic Blocks, a startup based in Oslo, has launched a beta version of a marketplace dedicated to the buying and selling of training data for humanoid robots. This platform became available on September 1, following months of development in collaboration with a select group of data suppliers and ea...
NVIDIA and LG Electronics have set a new benchmark in robot training data, aiming for 100,000 hours by year-end. This initiative was announced during a visit by NVIDIA's Senior Director of Omniverse and Robotics Marketing, Min-San Huang, to LG's Yangjae R&D Center in Seoul, following a strategic par...
Ropedia has announced the successful completion of a $22 million pre-Series A funding round, bringing its total funding to $30 million. The investment will be utilized to scale HOMIE, a lightweight, head-mounted device designed to capture first-person human movement and spatial context, which is ess...
Robots are becoming more visible in public spaces, captivating onlookers. However, they still lack the versatility needed for tasks in kitchens or factories, primarily due to a significant data bottleneck. Similar to human learning, robots acquire skills through experience, but the process of physic...
Tianqi Shares has accumulated nearly 20,000 hours of industrial data at its Jiangsu Embodied Intelligence Robot Industrial Data Collection and Training Center. This data supports the development of robotics applications in automotive production, with 36 primary and 163 secondary application scenario...
In 2026, China's humanoid robot shipments surpassed 40,000 units, capturing a 97% global market share. The Ministry of Industry and Information Technology forecasts that the total output for the year may exceed 100,000 units. However, a significant barrier remains in the form of data, as training th...
Rui Erman has launched a data training facility in Changzhou, equipped with 150 robots capable of generating 2GB of raw data per minute. This facility features two main areas: a basic motion training zone for fundamental robotic actions and a scenario application area designed for diverse operationa...
LG Electronics is enhancing its collaboration with Nvidia to expedite the creation of training data for humanoid robots. This initiative follows a memorandum of understanding signed by LG Group Chairman Koo Kwang-mo and Nvidia CEO Jensen Huang, aimed at expanding cooperation in physical AI and mobil...
MIT has introduced SceneSmith, a system utilizing AI agents to generate realistic 3D environments for robot training. This innovation addresses the challenge of providing diverse and rich simulation content, which is crucial for robots to learn effectively. By employing a vision-language model, Scen...
AI agents have developed the SceneSmith system, which creates realistic 3D environments such as kitchens and hotels for robot training. This innovative approach allows robots to simulate everyday tasks, enhancing their operational capabilities. The significance of this development lies in its poten...
MIT and the Toyota Research Institute have introduced SceneSmith, a system that utilizes AI agents to create realistic 3D environments for robot training. This innovation addresses the significant challenge of generating diverse simulation content, which is crucial for teaching robots various tasks...
The concept of 'robot training' often conjures images of individuals using VR headsets in labs, but this represents only a minor aspect of the field. The majority of job openings, totaling 1,196, are focused on operational roles that involve managing vehicles, fleets, and locations where robots int...
AGIBOT has announced the release of its Physical Reasoning dataset on October 22, focusing on the connection between planning and robot actions. The dataset aims to bridge the gap between high-level instructions and the specific actions required for robots to achieve tasks, such as restocking a hote...
Humanoid robots, designed to resemble human body shapes, are being developed to perform various chores and manual tasks. For these robots to be effective, they must reliably coordinate their movements, posture, and the way they interact with objects. This capability is essential for their successful...
AGIBOT has unveiled the Physical Reasoning dataset, marking the fourth release of its AGIBOT WORLD 2026 initiative. Announced on October 22, this dataset focuses on enabling robots to make decisions based on changing environments rather than merely executing complex movements. It emphasizes tasks su...
Generative Bionics is set to construct Europe's largest physical AI training facility for humanoid robots, utilizing AMD computing infrastructure. The initial 5,000-square-meter facility will be located in Italy and is expected to be operational by Q4 2027. This facility will be the first of three p...
Tesla CEO Elon Musk disclosed plans to establish the 'Optimus Academy' during the Q3 earnings call, aimed at facilitating AI training for the Optimus humanoid robot. He indicated that achieving the goal of producing 1 million robots annually may take over a year, necessitating the creation of 10,000...
The embodied intelligence industry is facing significant waste due to the need for retraining models whenever a new robot is introduced. As the number of robot types could reach tens of thousands, the industry is exploring concepts like 'one brain, multiple forms' to enable shared intelligence among...
Peking University and PrimeBot have developed HIL-UMI, a system that utilizes a handheld gripper to gather robot correction data when discrepancies arise between policy and human demonstrations. This innovative approach allows for more efficient data collection in robotic systems. The significance...
Dream Labs has made its public debut by launching the Dream-1 model, a no-data robot brain API, priced at $3 per hour. This innovative service allows developers to connect their dual-arm robots to the API, enabling real-time processing of visual data and command execution without the need for traini...
Physical Intelligence has introduced a beta version of its Pi API, enabling robotics developers to access its foundation models for fine-tuning and inference. This platform allows developers to upload robot data and fine-tune models on company-managed compute resources, streamlining the adaptation o...
Physical Intelligence has introduced a beta API designed for training and running its robot models, marking a significant advancement in providing intelligence for various machines. Announced on October 13, this service allows teams to upload demonstrations, train policies, and test them on hardware...
State Grid Tianjin Electric Power has recently opened two real-scenario training grounds for electric energy metering and cable tunnel inspection. These facilities aim to provide high-fidelity environments for robot training, algorithm iteration, and operational validation, enhancing the practical s...
In March 2026, Guanglun Intelligence, a physical-AI simulation infrastructure company based in Haidian, secured 1 billion yuan in A++/A+++ funding, marking its status as the world's first unicorn in the embodied data sector. The company does not manufacture robots; instead, it provides simulation da...
AWS has introduced its Physical AI Toolchain, designed to streamline the robot development process by integrating cloud resources with NVIDIA software. This toolchain aims to reduce the engineering time spent on connecting different stages of development, allowing teams to focus on improving robot c...
A robot training ground in Wuqing, Tianjin, has successfully hosted eight scene release conferences, showcasing advancements in robotics. This initiative highlights the growing importance of practical applications in robotics, moving beyond theoretical models to real-world scenarios. Observers sho...
Quantum Dynamics and KJ Logistics have initiated a 'Data Flywheel' experiment involving the deployment of robots in warehouse settings. This innovative approach aims to enhance operational efficiency and data utilization within logistics environments. The significance of this experiment lies in its...
Zhiyuan's data business is facing challenges as it operates at a rate of $20 per hour, which is insufficient to sustain the robotics industry. The company is exploring new avenues to enhance its offerings and address the growing demand for data in robotics. This situation highlights the broader iss...
Nanjing is set to host a robot conference that will not focus on performances but will instead emphasize the importance of welding training and the craftsman think tank. This shift in focus aims to enhance practical skills and knowledge in the robotics sector. The decision to prioritize welding tra...
Mecka has successfully raised $60 million in Series B funding, led by Sequoia Capital, to enhance its robotics data infrastructure and commercial deployments. The company specializes in capturing human demonstrations using proprietary multi-sensor hardware, converting this data into training signals...
Renesas Electronics Corporation has launched its first family of low-voltage 100V enhancement mode GaN-based discrete power transistors. This new lineup includes models RTP100E005G1FL, RTP100E2P6G1FL, RTP100E1P8G1FL-DSC, and RTP100E1P2G1FL-DSC, designed to deliver ultra-fast switching speeds and imp...
Innodata Inc. has inaugurated a new laboratory dedicated to capturing motion data essential for training humanoid robots. This facility aims to address the critical shortage of real-world interaction data that robotics teams face, as highlighted by CEO Rahul Singhal. The lab will also validate the p...
On September 28, the Luohu District of Shenzhen announced the launch of the first Luohu Cup, an integrated program for humanoid robot football competitions and training, held at Cuiyuan Middle School. This initiative, organized by the Luohu District Education Bureau and supported by Zhiyin Technolog...
Niantic Spatial has launched its Places Library, providing robotics developers with a catalog of 100 real environments for training and evaluation. The assets, available as USDZ files, include two representations for each environment: a Gaussian splat for visual appearance and a mesh for collision d...
On September 13, Zhao Yingwei's phone was flooded with messages from classmates and teachers after he was featured in a CCTV interview. Recently hired at an intelligent robotics company, Zhao shared his background in landscape architecture, prompting inquiries on how to enter the robotics field with...
Hygon Information Technology is preparing to launch a new chip aimed at physical-world applications, including robotics. This marks a significant shift from the company's current focus on data centers, as reported by Chinese media. The new chip, an iteration of the CPU1000 series, is designed to fu...
Recently, InTime Robotics, in collaboration with Fudan University and Shanghai Jiao Tong University, has officially open-sourced the HandEdit dataset and evaluation benchmark. This initiative focuses on first-person human-to-robot dexterous hand image editing, providing scalable data resources and s...
Watney has successfully raised $80 million to enhance its robotic solutions for data centers, focusing on the physical tasks necessary for increasing compute capacity. The funding round, announced on September 17, includes co-leads Valor Atreides AI Fund and Hummingbird Ventures, alongside returning...
Xiaomi has open-sourced the Xiaomi-Robotics-U0, an autoregressive embodied world foundation model featuring approximately 4 billion parameters and full-scale weight lines of around 38 billion. This release includes training and inference tools designed to enhance robotic applications. The significa...
Rhoda AI has reported enhancements in industrial manipulation capabilities through scaled web-video pretraining, as detailed in a study released on September 10. The research tested various model sizes, achieving completion rates of 3.7%, 65.0%, 75.3%, and 84.7% in under 100 seconds for tasks like u...
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
Daily robotics news, in-depth analysis, conference highlights, and discussions with professionals worldwide.
