AM-IA brings together news, research papers and applications relating to artificial intelligence published around the world into a single feed — and makes them available in your European language at the click of a button.
Scroll to browse · Click to filter
Tested on ex vivo pig colons, the system successfully completes 85.4 per cent of straight-line journeys and 70 per cent of 90-degree turns thanks to a policy that predicts sequences of actions.
A pilot feasibility study involving a humanoid robot operating in extended reality and using a pedal mechanism for construction tasks has shown a significant increase in the time required compared with manual labour.
The start-up has unveiled a ‘physical agent’ that carries out dozens of subtasks — washing, drying, folding and putting away — and decides for itself what to do next.
The computer vision platform aims to help models see, reason and act in three dimensions by publishing research, open-source models and testbeds.
The start-up, which does not manufacture any robots, has seen its valuation increase fivefold in just over a year and has exceeded $135 million in revenue and contracts with more than 30 clients.
After briefly topping the RoboDojo benchmark, OpenAI’s general-purpose model has improved the success rate of a block-placement task from 5 per cent to 95 per cent. Kai Williams plays down the significance of this.
Rather than letting its in-house actuators leak, Figure has used simulation to train its old robots to throw themselves into a Finnish arc furnace. The ingots that emerge are put up for sale.
Four fingers, a four-jointed opposable thumb and pressure sensors: Atlas’s new hand is designed to handle tools and practise in simulations.
Out of 103 proposals, the US Army has shortlisted five projects: a fully-fledged humanoid robot with ballistic protection, as well as power modules, tactile sensors and command software.
This vision-language-action model learns fine motor skills from first-person human videos, by representing the gestures in the camera’s coordinate system rather than mapping them onto the robot’s body.
Chinese researchers have presented an agent framework that coordinates a humanoid robot with the surgical team over long sequences, incorporating three overlapping layers of safety.
Anthropic has published a robot exposure index: at the historical rate of price falls, it would take forty years for robots to become competitive in 10 per cent of tasks.
Across four real-world tasks, vision-language-action policies trained solely on filmed human demonstrations from a first-person perspective perform on a par with those trained using teleoperation, whilst incurring lower data collection costs.
The framework adapts movements and scale to the robot’s body, reconstructs its states in simulation, and then the deployed model directly moves objects and reaches unknown targets.
The designer of inference chips for robots and drones went from 340 million to 1.45 billion in valuation, twelve months after a Series B of 85 million.
The bottleneck of visual imitation is the collection of clean videos. PRISM generates variants, reconstructs them into physically plausible trajectories, and deploys the policy without real robot adjustment.
Rather than a separate action expert, gestures become tokens added to the model's vocabulary. Grouped discrete broadcast decoding to hold real-time, and first places on the SIMPLE bench.
Opened on September 21 in Georgia, the RMAC center trains humanoids on real automotive logistics tasks. The assembly of components is not expected before 2030.
FedDRMan distills vision-language-action policies between heterogeneous robots through federated learning, and gains 11.6 points over the best competing federated method.
Collision with a Nissan Murano in Las Vegas on September 25, a hospital driver, cause not established — and the limit of 100 vehicles lifted on the same day.