RoboSignal Daily · 2026-10-17 · Saturday
ROBOSIGNAL
Open-sourcing ASPIRE V1.0: Agentic Robotics Framework with Real Robot Skill Discovery
Jim Fan's team at X Robotics open-sources ASPIRE V1.0, an agentic robotics framework with real robot skill discovery and self-improvement capabilities. The framework was tested on real robots and benchmarks, achieving ~99% performance overnight. The release includes updates on model performance and cost efficiency in robot tasks.
Research
Open-sourcing ASPIRE V1.0: Agentic Robotics Framework with Real Robot Skill Discovery
Jim Fan's team at X Robotics open-sources ASPIRE V1.0, an agentic robotics framework with real robot skill discovery and self-improvement capabilities. The framework was tested on real robots and benchmarks, achieving ~99% performance overnight. The release includes updates on model performance and cost efficiency in robot tasks.
ETH Zurich scientists turn WUJI robotic hand into self-propelled mobile manipulator
ETH Zurich researchers have developed a system where a WUJI robotic hand can crawl, recover from falls, and perform manipulation tasks without a separate mobility system. Using reinforcement learning, the hand learns to support its own weight and move using its fingers. The work was presented in a paper titled 'Fingers as Legs: Learning Self-Supported Locomotion and Manipulation with an Anthropomorphic Hand.'
Revisiting the Forklift as a Robot: JieXiang's New Approach to Warehouse Automation
JieXiang has redefined the forklift as a robot by eliminating human-centric features like driver seats and protective structures, reducing redundancy in structure, energy, and cost. The company's approach focuses on original robot design principles, modular architecture, and system-level coordination to simplify the deployment of automated storage systems.
Observing the Industrial Fair: Embodied Intelligence Moves into Factories, What Capabilities is Faol Robotics Filling?
At the Industrial Fair, the integration of embodied intelligence with robots is bringing about subtle changes. The robots on display are not only required to demonstrate mobility but also to address real-world challenges such as identifying unfamiliar workpieces, adjusting operations, moving around, and force control. Faol Robotics showcased multiple technical paths, including the AIRLab intelligent welding solution, the ART series fully joint-force-controlled humanoid robotic arms, and the BW10-Lite industrial-grade embodied robot, aiming to enhance robots' adaptability to complex industrial tasks.
After Collaborative Robot Deliveries Surpass 50,000 Units, This Company Brings Embodied Intelligence into Factories
At the 2026 China International Industrial Robot Expo, ABB Intelligent Robotics showcased its progress in deploying flexible manufacturing solutions, including a complete automotive production line with G-series and iS-series collaborative robots, as well as a new dual-arm robot and standardized welding workstations. The 50,000-unit delivery milestone marks a key step in its industrial deployment strategy.
RoboAtlas: Contextual Active SLAM
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.
Chinese Robot Plays Badminton for 1452 Consecutive Shots. Sets Guinness World Record
The badminton robot developed by Shenchendai Dong completed 1452 consecutive shots in Shaoxing, earning a Guinness World Record. The robot is based on a self-developed general-purpose physical motion platform, featuring real-time perception, trajectory prediction, and coordinated control capabilities, showcasing the engineering strength of robots in dynamic environments. The article analyzes its technical approach and potential applications, including logistics, warehousing, and home service robots.
Revisiting Embodied Chain-of-Thought for Generalizable Robot Manipulation
This research explores embodied chain-of-thought (CoT) for robotic control, introducing ERVLA, a vision-language-action model that effectively integrates linguistic reasoning into robot manipulation. ERVLA uses a CoT-dropout strategy during training and predicts actions directly during inference, achieving 86.9% success on LIBERO-Plus and 53.2% on VLABench, with superior performance in out-of-distribution settings.
Briefs
- Spark-0: Scaling Spatially Aligned Tactile Pretraining for Robot Dexterous ManipulationarXiv Robotics — research abstracts