Companies & Labs
Robotics and embodied AI developments associated with Unitree. Generic AI/company coverage is excluded.
24 selected reportsLast 30 days 21 recordsAll coverage 175 records
JiZhiJia invested 50 million RMB in ShenZhi Robotics, acquiring a 20% stake. ShenZhi unveiled its self-developed embodied base model SE-1 and two robots at the 2026 Industrial Fair: Sculptor for multi-task heavy-duty logistics operations, and Edge for industrial precision manipulation.
SuperMimic uses a single differentiable world model to scale humanoid whole-body control across motions, achieving faster training and lower computational costs. It outperforms existing baselines in both efficiency and accuracy, with minimal failure rates on benchmark motions.
This paper presents LocoPilot, a single-operator bilateral teleoperation framework for loco-manipulation policy learning. It enables direct capture of robot-specific signals like proprioception and body-mounted observations without motion-capture instrumentation. The system integrates a 12-DoF quadruped with a 6-DoF manipulator, using two handheld controllers and a follower arm for force feedback. A ROS~2 pipeline records synchronized multimodal data and a unified 11D action space, validated through deployment on a physical robot without retargeting or additional inverse kinematics.
The article discusses the decline in robot performance rental prices in China, driven by increased supply and competition, while highlighting the emergence of professional, IP-driven robot performances in international events like 'America's Got Talent.' It emphasizes the transition from hardware-centric to content-driven models, with entertainment and cultural tourism as key growth areas for embodied AI.
The article discusses the decline in robot performance rental prices in China due to increased supply and competition, while highlighting the international success of Chinese robots in global entertainment events like 'America's Got Talent.' It underscores the transition from hardware-centric to content-driven robotics performance.
The article discusses the decline in robot rental prices in China, driven by increased supply and competition, while highlighting the growing role of robotics in entertainment and cultural events. Examples include the use of humanoid robots in international performances and theme parks, which provide real-world interaction data for training and development.
The U.S. is implementing two simultaneous regulatory actions against the Chinese robotics industry: a FCC ban on new foreign-made advanced robotics equipment and a $200,000 fine on a U.S. firm for failing to report an investment in a Chinese embodied intelligence company. These measures target the industry's core competitive advantage—iteration speed—by restricting both product imports and capital flows, signaling a preemptive strategy to curb China's influence in the physical-intelligence domain.
A research team from Shanghai Chuangzhi College, Yuance Future, Hong Kong University, and NVIDIA has developed PingpongDuo, a system that enables two Unitree G1 humanoid robots to perform table tennis doubles. The system combines perception, team coordination, and whole-body motion control, using reinforcement learning for planning and risk assessment. Real-world testing showed improved performance over traditional rule-based strategies, with the robots achieving 18 consecutive alternating returns in a human-robot doubles match.
RoboAtlas is a contextual Active SLAM framework that adaptively balances geometric exploration and semantic reasoning using a scalable 3D semantic mapping system. Evaluated in simulation and on a Unitree Go2 robot in large-scale real-world environments, it achieves a 100% task success rate and state-of-the-art performance on the GOAT-Bench 'Val Unseen' benchmark, with success rates of 90.6% using GPT-4o and 88.8% using Qwen2.5-VL-7B.
This research presents LEGS, a hybrid simulator that combines a mesh foreground with a photorealistic 3D Gaussian Splatting background to train vision-language-action (VLA) policies for humanoid loco-manipulation without human teleoperation. The method uses a procedural motion-primitive generator to synthesize labeled demonstrations and a two-stage color calibration to align renders with the deployment camera. Policies trained on LEGS data match or exceed those trained on human teleoperation demos and outperform mesh-only baselines. The approach enables cost-effective re-rendering of demonstrations for new scenes and objects.
This paper introduces ASENA, an embodied agent system that enables general-purpose coding agents to program humanoid navigation, manipulation, and interaction. ASENA-VLN, a 4B monocular navigation policy, achieves state-of-the-art success rates on R2R and RxR benchmarks. Integration with coding agents improves success rates and reduces execution time, while persistent workspace evolution enhances performance through simulator feedback.
In 2026, Embodied AI becomes a hot topic in the AI field. Liang Wenfeng, founder of DeepSeek, has recently intensively engaged with multiple hard-tech startup teams to promote investment, joint R&D, and the establishment of in-house teams. The article analyzes the technical challenges and practical application scenarios of Embodied AI, including the deployment of robots in factories, elderly communities, and educational sectors, and discusses its profound impact on the AI industry.
This research presents Long-WAM, a model-system framework that scales the context of causal world-action models under real-time control constraints. By using autoregressive pretraining, the model improves task success rates on benchmarks like RoboCasa GR-1 and LIBERO-Long. It also supports real-time deployment on hardware platforms, enabling dynamic and long-horizon manipulation tasks.
In 2026, the robot industry shifted from hardware competition to a contest of 'brain' capabilities. The article analyzed the technical stack of 'robot brains,' including perception, cognition, planning, and control layers, and discussed the importance of data loops. The article focused on two companies, Yu Shi Technology and Xian Gong Intelligence, representing the 'body-centric' and 'brain-centric' approaches respectively, exploring their deployment strategies and data accumulation methods in real-world scenarios.
The article discusses the shift in the robotics industry from speculative valuations to practical deployment in manufacturing environments. It highlights the challenges of integrating robots into production lines, including navigation, vision systems, communication protocols, and safety considerations. The piece emphasizes the importance of real-world performance metrics like cycle time, yield, and throughput, and the need for robotics companies to focus on delivering tangible value through industrial applications rather than speculative hype.
At the 2026 World Robotics Conference, Sergey Levine pointed out that the robotics industry is undergoing a shift from controlling the body to developing decision-making capabilities. He emphasized that robots need to combine prior knowledge with real-world experience to achieve true intelligence. Although robots are still in the foundational technology stage, through data flywheels and scaling, they may eventually enable broader applications.
MW Company launched a residential project integrating mechanical arms into ceiling tracks, with robots to be installed as late as 2028. The residential units are priced between 200 million and 400 million yen, adopting a vertical integration model that combines housing, software, and robotics in a single development. The company aims to achieve home intelligence through continuous upgrades and maintenance systems, with the goal of making homes with robots a standard option.
The article explores the rapid expansion of embodied intelligence data collection centers in China, driven by both policy and market forces. It also reveals issues with some companies creating false revenue through a 'equipment sale—data repurchase' model. The article further analyzes the technical challenges of centralized data collection and the industry's shift toward distributed data collection models.
The Stanford team's HomeBody project demonstrates how GPT-6 Astra controls the Unitree G1 robot to complete kitchen cleaning tasks through a three-step process: spatial exploration, digital twin simulation, and task execution. The system uses pre-trained models and a skill library to perform complex operations in new environments without additional training.
The article explores the issue of data scarcity in the field of embodied intelligence, pointing out that although there are already a large number of data collection centers, high-quality and reusable data remains severely insufficient. The article analyzes structural contradictions such as high data collection costs, inconsistent quality, fragmented formats, and lack of cross-ontology reusability. It introduces Wu Wen Tech's solution of building a data foundation through a Real2Sim2Real closed-loop system, including large-scale data collection, automated annotation, and simulation training technologies, ultimately forming a data-driven flywheel to promote the development of embodied intelligence.
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).