New Technology– category –
-
HKUST(GZ): 35.7 Nm/kg torque density robot actuator specs
EYOU China Overview This article details four key engineering methods to enhance robot actuator power density: coaxial integration, miniaturized reducers, topology-optimized housings, and optimized motor slot filling. Notably, a coaxial ... -
MIT: Paper-thin swimming robot powered by live muscle cells
SciTechDaily USA Overview MIT engineers have created a groundbreaking paper-thin swimming robot driven by a layer of living muscle tissue thinner than a human hair. This biohybrid robot utilizes light-controlled living muscle cells for l... -
Dongguk University: Battery-free wearable sensing device explained
Robotics & Automation News South Korea Overview Researchers at Dongguk University have developed a flexible, battery-free wearable electronic device capable of harvesting energy from human movement to power neuromorphic sensing and learn... -
Zhejiang University: 200V artificial muscle specs for soft robots
Sina Finance China Overview A team led by Professor Shi Ye at Zhejiang University has developed a novel artificial muscle that operates with a mere 200 volts, paving the way for soft robots to operate without external power cables. This ... -
Silicon Li-ion Batteries: 30-40% energy density boost for drones
Ayaa Unknown Overview New silicon anode Li-ion batteries for military drones deliver a 30-40% increase in energy density compared to conventional graphite cells, reaching 350-450+ Wh/kg. This advancement significantly extends the enduran... -
XonoAI Reports 99.8% Object Detection Accuracy in Autonomous Vehicles with 4D Millimeter-Wave Radar and End-to-End Foundation Models
XonoAI USA Overview XonoAI's latest report highlights a significant shift towards end-to-end foundation models in autonomous driving, moving beyond modular perception stacks. A key advancement involves a cross-modal attention network fus... -
RCVLA Model Enhances Autonomous Driving Planning Accuracy and Safety with 4D Radar and Semantic Reasoning
arXiv International Overview This research proposes 'RCVLA,' a new radar-camera Vision-Language-Action (VLA) framework leveraging 4D radar data. RCVLA significantly improves autonomous driving system planning accuracy and safety through ... -
Yann LeCun’s AD-E2E-JEPA Accelerates Autonomous Driving Trajectory Planning 100x Without Policy Learning
AlphaSignal USA Overview A research team including Yann LeCun has proposed "AD-E2E-JEPA," a self-supervised world model that determines driving trajectories without explicit policy learning. This model operates approximately 100 times fa... -
iDriveVLA Overcomes Multimodal Autonomous Driving Challenges, Achieves New SOTA of 94.95 PDMS on NAVSIM v1 Leaderboard
arXiv Unknown Overview As multimodal planning addresses ambiguous scenarios in autonomous driving, the novel iDriveVLA framework has overcome generation-evaluation asymmetry, achieving a new state-of-the-art performance of 94.95 PDMS on ... -
FAA Proposes Part 108 BVLOS Rules to Revolutionize Power Line & Substation Inspections with AI Drones
Reboot Hub USA Overview The FAA's proposed Part 108 BVLOS (Beyond Visual Line of Sight) rules aim to designate airspace within 50 feet of power lines and substations as "shielded areas," enabling routine drone operations for infrastructu...
