New Technology– category –
-
GAMs explained: 3D Grounded Action Models boost accuracy 61%
arXiv International Overview A new arXiv paper introduces Grounded Action Models (GAMs), a novel paradigm for robot foundation models built on 3D grounding, achieving up to 61% average success rates. GAMs can be conditioned using languag... -
DexTouch-WM explained: Human haptics for robot manipulation
arXiv International Overview A novel framework, DexTouch-WM, successfully learns predictive models for contact-rich dexterous robot manipulation from scalable human tactile data. This model can jointly forecast future RGB observations an... -
Rincell RC50T: Tabless 21700 Li-Ion cell specs and 265 Wh/kg
DRONELIFE USA Overview U.S. battery cell manufacturer Rincell unveiled its NDAA-compliant tabless 21700 lithium-ion cell, 'RC50T,' at DroneX. Designed specifically for drones, this new cell features 100A continuous discharge, 150A pulse ... -
University of Stuttgart: Ceramic microscrolls specs for 2026
PNAS (Proceedings of the National Academy of Sciences) Germany Overview Researchers at the University of Stuttgart have developed magnetically actuated ceramic microscrolls with mechanically stored elastic energy, pioneering a new class ... -
Pony.ai and Kodiak Advance Autonomous Driving by Integrating Large-Scale AI Models with Robust Simulation for Complex Challenges
New Market Pitch USA Overview Pony.ai and Kodiak are tackling the core challenges of autonomous vehicles—geographic generalization, long-tail safety events, scalable economics, operational complexity, and regulation—through advanced AI m... -
Multi-Modal Perception Pipeline for Autonomous Racing Achieves Robust Object Tracking with Latency Compensation
arXiv Unknown Overview A new multi-modal late-fusion perception pipeline, designed for autonomous racing, integrates independent detections from cameras, LiDAR, and radar to provide timely and robust state estimation of surrounding vehic... -
DRS-VPT Achieves State-of-the-Art Performance in Direct Image-to-Scan Relocalization, Integrating Autonomous Driving LiDAR Calibration
arXiv Unknown Overview The novel DRS-VPT feed-forward transformer architecture has achieved state-of-the-art performance in direct image-to-scan relocalization. This model accurately predicts scan poses and point cloud maps from query im... -
Formal Verification Uncovers Hidden Flaws in Autonomous Driving AI Missed by CARLA Simulation
arXiv Unknown Overview A new study demonstrates that formal verification can identify critical failures in AI-based autonomous driving systems that are undetectable through conventional CARLA simulations. By testing an end-to-end steerin... -
Motional Open-Sources “nuReasoning” Dataset, World’s First Reasoning-Centric ‘Long-Tail’ Data for Enhancing Human-Like Reasoning in Autonomous Vehicles
Zhongjin Online China Overview Motional has announced nuReasoning, the world's first open-source, reasoning-centric 'long-tail' scenario dataset aimed at helping autonomous vehicles (AVs) develop human-like reasoning and decision-making ... -
MicroVision to Unveil Next-Gen Modular LiDAR Perception Platform at IAA Transportation 2026, Offering Versatile Integration for Diverse Applications
Accesswire (via MicroVision, Inc.) USA Overview MicroVision, Inc. announced plans to preview its next-generation "MicroVision Perception Platform™" at IAA Transportation 2026. This modular platform integrates the company's LiDAR hardware...
