Inside China’s robotics revolution
The long read: How close are we to the sci-fi vision of autonomous humanoid robots? I visited 11 companies in five Chinese cities to find out
📍 The Guardian🕐 28 days ago
Is it possible to develop AI without the US?
Tech giants Alphabet, Amazon, Microsoft and Meta to collectively invest $600bn on artificial intelligence this year
📍 theguardian.com🕐 65 days ago
Hail our new robot overlords! Amazon warehouse tour offers glimpse of future
At its new Stone Mountain, Georgia, facility, Roomba-like robots shuffle between stacks, another adds shipping labels while another arranges packages in pallets
📍 theguardian.com🕐 69 days ago
Tesla discontinues Model X and S vehicles as Elon Musk pivots to robotics
High hopes for Optimus robot help company beat forecasts despite yearly revenue decline and flailing car business
📍 theguardian.com🕐 77 days ago
ETH Zurich Reactive Walking Controller Handles 30 cm Step Heights Without Falling
A new model-predictive control framework from ETH Zurich's Robotic Systems Lab allows humanoid robots to traverse uneven outdoor terrain at near-human walking speeds. The 1 kHz controller handles unexpected step heights up to 30 cm without falling. Testing on ANYmal D showed a 98.7% success rate on a challenging outdoor obstacle course.
📍 Robotics: Science and Systems🕐 3 days ago
New Vestibular Sensor Module Cuts Humanoid Balance Recovery Time by Half
A biologically inspired vestibular sensor module developed at Imperial College London reduces the time humanoid robots take to recover balance after unexpected pushes by 52%. The module mimics the inner ear's semicircular canals, providing high-bandwidth angular acceleration estimates to the balance controller. Integration trials with three commercial robots showed zero falls during standardized push-recovery tests.
📍 Nature Robotics🕐 3 days ago
Proactive Safety Architecture Enables Robots to Predict and Avoid Human Collisions
Researchers at TU Munich published a proactive collision avoidance framework for humanoid robots operating in human-shared workspaces. The system predicts human trajectory intentions 1.2 seconds ahead and adjusts robot motion accordingly, reducing near-miss events by 87% in a factory simulation. The framework is compatible with existing robot operating system stacks.
📍 Science Robotics🕐 4 days ago