Technical Directions
Robotics coverage relating to robot policies.
335 selected reportsLast 30 days 289 recordsAll coverage 1,886 records
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.
AGIBOT Robotics has deployed 100 units of the Lingxi X2 humanoid robot in 100 retail stores of Aishida, achieving a 39% increase in store sales. The robots, named 'Aibao,' engage customers by understanding their cooking needs and guiding them toward suitable products, while also seamlessly transferring conversations to human staff when necessary. This marks a milestone in the deployment of embodied AI in real-world retail environments.
The article analyzes China's leading position in the field of humanoid robot patents, noting that as of the first half of 2026, China holds 60.8% of relevant patents, significantly surpassing the United States and South Korea. The article examines the evolution of patent strategies, the trend of technological shift from hardware to software, and the progress made by Chinese companies in patent quality, application scenarios, and global layout. It also points out that patent quantity does not equate to comprehensive technological leadership, emphasizing the challenges of manufacturing costs and quality, as well as the dual role of patents in industrial competition.
Robotic Systems Lab@leggedroboticsSignalEditorial score6262
Carnegie Mellon University and Siemens developed PRISM, a method to assess the safe stoppability of humanoid robots. The approach uses simulation to define safe stopping conditions, enabling proactive safety monitoring and supporting certification of fail-safe behaviors. The research highlights the importance of context in determining when a robot can be safely stopped, especially in environments with people or obstacles.
Carnegie Mellon University researchers have developed LAMP, a system that allows multiple robots to work together to move objects through crowded spaces. LAMP combines learned models with search-based planning to enable efficient navigation and coordination, with successful testing in complex environments and a demonstration at the 2026 IROS conference.
Animesh Garg@animesh_gargSignalEditorial score8585Introducing FLUX 3 Action. An open weights 7B World Action Model that achieves first place on the RoboLab benchmark. It outperforms the previous best open model by 6.1 percentage points while using 56% fewer parameters and running up to 3.95x faster. FLUX 3 Action removes the usual trade-off between world action model performance and VLA speed: it still predicts video and actions together, but plans more than twice as far ahead and runs faster per second of robot motion than the strongest open VLA. Teams can fine-tune FLUX 3 Action on their own demonstrations to create policies for a particular robot and task. Together with @nvidia, we also integrated FLUX 3 Action natively into @huggingface's LeRobot, with fine-tuning recipes included and edge deployment on NVIDIA Jetson. Beyond robotics, we’re also seeing promising results training task-specific policies for acting in simulated environments like gaming, controlling a vehicle, computer use, and wherever else a model needs to understand a visual environment and then choose what to do next. FLUX 3 Action builds on the same image, video, and audio pretraining as FLUX 3, but uses a smaller architecture designed for practical deployment. In midtraining, we trained the model to predict actions and future frames together. We’re releasing the weights, code, fine-tuning recipe, benchmarks, and reproducible examples so researchers and developers can build on the model with their own robots, environments, and tasks (see below).
China is pursuing a strategic roadmap through its Five-Year Plans to dominate the global robotics industry, focusing on embodied AI and humanoid robotics. The 13th and 14th FYPs laid the foundation for industrial automation, while the 15th FYP emphasizes embodied intelligence, with state support for R&D and mass deployment of humanoids. Despite achievements in market share and production scale, challenges remain in precision hardware and semiconductor self-sufficiency, with risks of overcapacity and market volatility.
Robotic Systems Lab@leggedroboticsSignalEditorial score6363
Researchers at the University of Michigan and UCLA have developed a new propulsion mechanism for small robots using low-power twisting of elastic rods. This method enables hopping and swimming motions by utilizing snapping mechanics. The mechanism works on challenging surfaces such as slippery cloth and grass-covered ground, producing frog-like motion. The study, published in Science Advances, was co-led by Khalid Jawed and Xiaonan (Sean) Huang, and demonstrates the potential for miniature robots to maneuver over uneven or slippery terrain.
MIT researchers have developed a reconfigurable robotic lab that autonomously assembles, tunes, and dismantles optical experiments. The system uses robotic arms, 3D-printed component housings, and a cloud-based interface to enable remote operation. It can build and fine-tune a laser cavity in under 30 minutes, demonstrating the potential for fully automated optical experiments.
Researchers at ETH Zurich have developed a robotic platform with a magnetically controllable endoscope to improve the safety and precision of fetal surgery for twin pregnancies with shared placentas. The system uses magnetic fields to bend the endoscope, enabling better access to blood vessels, and combines panoramic imaging with real-time navigation. The technology has been tested in an animal model and is now being refined for human clinical use.
MIT researchers have developed a new technique that helps generative AI models meet strict safety and task-specific constraints without sacrificing output quality. The method, called HardFlow, reformulates constraint satisfaction as a trajectory-optim, allowing models to explore more freely during generation while ensuring final compliance with hard constraints. It is tested on robotics, control, and computer vision tasks, consistently outperforming existing methods in constraint satisfaction and solution quality.
This article explores how tactile data is helping robots improve their dexterous manipulation skills. Researchers are creating large tactile datasets and models that can use tactile feedback to enhance robot performance in tasks like folding laundry or turning keys. Challenges include the difficulty of integrating tactile data with vision-based models and the need for more diverse and scalable datasets to achieve significant improvements in robot dexterity.
Researchers at the University of Queensland and the University of New South Wales have developed 'paraborgs'—cyborg cockroaches equipped with drug-delivery systems for search-and-rescue missions. These insects can navigate through rubble, locate survivors, and administer medication using remote control and spring-loaded needles. While the technology shows promise with high success rates in tests using silicone and pig skin, human trials have not yet been conducted.
Researchers at the University of Queensland have developed 'Paraborgs'—cyborg cockroaches equipped with needles and cameras to assist in search-and-rescue missions. These insects, grafted with electronic interfaces, can navigate rubble, deliver drugs, and operate in teams. The study addresses challenges in autonomy, communication, and human interaction with the technology.
Skild AI introduces S1, a robotic foundation model that leverages in-context learning to execute complex, long-horizon tasks without post-training. This marks a significant shift from traditional fine-tuning approaches, enabling rapid deployment and reducing data requirements. The model demonstrates strong performance on unseen tasks, including plant potting, pancake cooking, and kit assembly, and shows robustness to perturbations and common-sense reasoning.
Researchers at MIT have developed FloatForm, a system of robotic tiles that can autonomously assemble into floating structures. Inspired by ant colony behavior, the system uses magnetic latches, sensors, and thrusters to reconfigure dynamically, addressing challenges like hydrodynamics and slipstreams. Potential applications include emergency response, public space, and infrastructure on water.
Google DeepMind has launched Gemini Robotics ER 2, a new model that enhances robotics through video understanding, task orchestration, and multi-robot collaboration. It enables robots to perform complex tasks, adapt in real-time, and work together in shared environments. The model also improves safety and spatial reasoning capabilities, with performance benchmarks showing significant gains over previous versions.
NEURA Robotics presents a seven-stage pipeline for training robots, from defining use cases and capturing real-world data to model training, optimization, and deployment. The process emphasizes physical AI, multimodal data collection, and iterative refinement to ensure reliable, scalable robotic capabilities.
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).