Skip to content
2026-10-24 · SaturdayDaily briefing

RoboSignal Daily · 2026-10-24 · Saturday

ROBOSIGNAL

Issue 25242026-10-24Saturday
12events10sources1first-party announcementsAbout 15 minutes to read
Lead story

AGIBOT Open-Sources AGIB T WORLD Release 4: 'Physical Reasoning' Dataset Enabling Robots to Reason Before Acting

AGIBOT has released the fourth version of its AGIBOT WORLD dataset, focusing on 'Physical Reasoning' to enable robots to make decisions before acting. The dataset covers over 10 categories of logical reasoning scenarios, supporting both high-level decision-making models and low-level action policies. It includes annotations linking task goals to execution processes, aiding in the training of embodied agents for real-world tasks.

01

Research

AgiBot — Official newsFirst-party

AGIBOT Open-Sources AGIB T WORLD Release 4: 'Physical Reasoning' Dataset Enabling Robots to Reason Before Acting

AGIBOT has released the fourth version of its AGIBOT WORLD dataset, focusing on 'Physical Reasoning' to enable robots to make decisions before acting. The dataset covers over 10 categories of logical reasoning scenarios, supporting both high-level decision-making models and low-level action policies. It includes annotations linking task goals to execution processes, aiding in the training of embodied agents for real-world tasks.

Surgical robotics — News discovery

Surgical AI Learns to Adapt When Smoke and Darkness Blur the Robot’s View

A research team at Chengdu University of Information Technology has developed LiteST-TTA, a framework for real-time adaptation of surgical segmentation models during minimally invasive procedures. The model uses uncertainty estimation and temporal consistency to adjust predictions in response to visual degradation like smoke and low illumination. Evaluated on the EndoVis 2017 dataset, LiteST-TTA achieved a Dice similarity coefficient of 0.8879, outperforming the fixed-source baseline in severe low-light conditions. However, the study highlights that adaptation is not universally beneficial and depends on the type and severity of visual degradation.

Chinese robotics — Research news discovery

ViGAR: First World-Action Model for Full-Scale Humanoid Robots

ViGAR, developed by Zhiyuan and Peking University, is the first complete world-action model (WAM) for full-scale humanoid bipedal robots. It has been validated on the real-world AgiBot A2 platform and achieved strong performance on the RoboTwin 2.0 benchmark. The model introduces a hierarchical architecture with subgoal planning and execution, enabling long-horizon task completion through explicit reasoning and context learning.

Robotics — Robot learning news discovery

NVIDIA Expands Isaac Tools for Humanoid Robot Development

NVIDIA has expanded its Isaac open robotics development platform with a humanoid reference platform, new datasets, and runtime frameworks to support robot learning workflows from simulation through deployment. The Agile ONE SN platform integrates simulation, training, and deployment tools, offering a full-stack solution for humanoid robotics research and development.

Chinese robotics — Research news discovery

ViGAR: The First World-Action Model for Full-Sized Humanoid Robots

In October 2026, Zhuyuan Robotics collaborated with the DA Group at Peking University and CocoMatrix to release ViGAR, the first complete world-action model (WAM) for full-sized humanoid bipedal robots. ViGAR has been validated on the real-world AgiBot A2 platform and achieved strong results on the RoboTwin 2.0 benchmark, ranking third in the real-time leaderboard. The model introduces a hierarchical architecture with a subgoal planner and executor, enabling long-horizon task execution through explicit visual subgoal reasoning. This approach allows the robot to plan and execute complex, multi-step tasks in dynamic environments.

Chinese robotics — Company news discovery

Robot Insurance Pricing Challenges: From Data Gaps to Risk Quantification

The article discusses the difficulties in pricing robot insurance, including the lack of historical data, the high risk of accidents, and the complexity of liability. It highlights the first insurance claim in China, the development of insurance models by major insurers, and the industry's efforts to create standardized risk assessment and liability frameworks for robot-related incidents.

Robotics & Automation News

ETH Scientists Develop Biohybrid Actuators with Lab-Grown Muscles for Visible Movement

Scientists at ETH Zurich have created robot actuators using living skeletal muscle tissue grown from mouse cells, integrated with 3D-printed hydrogel scaffolds. The biohybrid actuators, designed through computer simulations, can flex visibly and perform actions like jumping and swimming. While the system shows promise, challenges remain in long-term operation due to nutrient requirements and tissue variability.

Chinese robotics — Hardware and sensing

SoftBank's Gamble on Physical AI: A Billion-Dollar Bet on Robotics

SoftBank's CEO Masayoshi Son is investing heavily in robotics through acquisitions of hardware firms like ABB and 1X, AI models from OpenAI and Skild AI, and real-world scenarios such as security and cleaning services. Despite market skepticism and a 10% stock drop, the strategy aims to create a closed-loop physical AI ecosystem, though challenges remain in aligning model progress with hardware deployment.

RoboSpeak — WeChat

SoftBank's Robotics Ambition: A Billion-Dollar Bet on Physical AI

SoftBank is aggressively investing in robotics by acquiring industrial robot companies like ABB, humanoid firms like 1X, and AI startups such as Skild AI. The company is also building a domestic AI model supply chain and acquiring service providers like Allied Universal to create real-world deployment scenarios. Despite market skepticism and past failures, SoftBank aims to establish a complete physical AI ecosystem through hardware, software, and scenario integration.

Chinese robotics — Robot learning research

Physical AI and Embodied Intelligence: Industry Signals Point Toward a Critical Inflection Point

The article discusses the convergence of physical AI and embodied intelligence, with key industry events and product launches signaling a shift toward standardized robot configurations and model-driven production. The release of H2-X and GPT-6Astra demonstrates progress in direct model control and physical interaction, while policy support and market data suggest the industry is approaching a critical mass.

Robotics — Chinese web discovery

Annual Production of 1 Million Joint Modules. Ant Factory Collaborates with Quanzhibo and Lingxin Qiaoshou to Target Core Embodied Intelligence Track

Ant Factory, in partnership with Quanzhibo and Lingxin Qiaoshou, has established a joint venture to produce 1 million joint modules annually, targeting the core components needed for embodied intelligence robots in the Beijing-Hebei region. The collaboration aims to address the growing demand for joint modules in the humanoid robot industry, leveraging each company's expertise in manufacturing, joint module development, and end-effector integration.

02

Robots & Hardware

Robotics — Robot learning news discovery

RoboParty RP1 (ROBOTO 01) – A full-stack open-source humanoid robot with up to 30 degrees of freedom

RoboParty has launched the RP1 (ROBOTO 01), an open-source humanoid robot with up to 30 degrees of freedom, designed for robotics research and embodied AI development. The robot comes in two configurations, with detailed hardware specifications, including Rockchip and NVIDIA processors, custom joint actuators, and a range of sensors. The platform is supported by open-source repositories for firmware, simulation, and deployment.

03

Briefs

End of issue

RoboSignal Daily is automatically compiled from eligible stored coverage; each item links to its original source · Briefing archive

Timezone · UTC

Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).