Researchers studied how reinforcement learning enabled a robot arm equipped with a parallel jaw gripper and tactile sensors to insert unknown objects into a hole.

Multi-domain operations (MDO) the Army’s future operating concept, requires autonomous agents with learning components to operate alongside humans. New Army research reduces the unpredictability of current training reinforcement learning policies so that they are more practically applicable to physical systems, especially ground robots. These learning components will permit autonomous agents to reason and adapt to…

NVIDIA said its Isaac Gym simulation software is designed to provide end-to-end reinforcement learning for robotics developers.

Army researchers developed a reinforcement learning approach that allows swarms of unmanned aerial and ground vehicles to optimally accomplish various missions while minimizing performance uncertainty. Swarming is a method of operations where multiple autonomous systems act as a cohesive unit by actively coordinating their actions. Army researchers said future multi-domain battles will require swarms of…

Intel AI Lab is working with researchers at Oregon State, Stanford, and UC San Diego on machine learning approaches that could help robots interact with dynamic environments. They include a combined approach to reinforcement learning and PartNet, a massive dataset of 3D objects with annotated components.

Quick demonstration of a converged policy using ROS2Learn framework and the gym-gazebo2 toolkit. We execute a deterministic run and also use settings that replicate a real behavior of the robot.   The first gym-gazebo was a successful proof of concept, which is being used by multiple research laboratories and many users of the robotics community.…

Industrial robots deployed today across various industries are mostly doing repetitive tasks. The overall task performance hinges on the accuracy of their controllers to track predefined trajectories. The ability of robots to handle unconstructed complex environments is limited in today’s manufacturing. Two examples are flexible picking of previously not encountered objects or the insertion of…

The sun may be setting on what David Letterman would call “Stupid Robot Tricks,” as intelligent machines are beginning to surpass humans in a wide variety of manual and intellectual pursuits. In March 2016, Google’s DeepMind software program AlphaGo defeated the reining Go champion, Lee Sedol. Go, a Chinese game that originated more than 3,000…

With her fun, bubbly personality, ABii is currently making waves in classrooms around the world, building both core skills mastery and confidence.

Wayve, a UK startup founded in 2017, raised a $20 million Series A round to launch a pilot fleet of self-driving Jaguar I-Pace electric SUVs in central London. Wayve believes the engineering challenges of self-driving cars will be solved by better artificial intelligence, not by more sensors and hand-coded rules. Wayve trains its autonomous driving…

Unity Technologies released the open beta version of its Unity Machine Learning Agents, an artificial intelligence toolkit developers and researchers can use to virtually train agents —whether video game characters, autonomous vehicles or robots. “Machine learning is a disruptive technology that is important to all types of developers and researchers to make their games or…

Intrinsic has been exploring how automated perception, deep learning, reinforcement learning, motion planning, force control, and simulation can be combined to make industrial robots more flexible.

In this episode of The Robot Report Podcast, we talk to Jonathan Hurst, Co-Founder and CTO of Agility Robotics, which is developing legged robots. We discuss how passive dynamics has steered the direction of Agility’s robots, the role of simulation and reinforcement learning, the fourth version of Digit currently being developed, comparisons to Boston Dynamics, and much more.

Researchers from Columbia Engineering have designed a robot hand that joins an advanced sense of touch with motor-learning algorithms.

MIT’s DribbleBot can maneuver soccer balls on landscapes like sand, gravel, mud and snow and get up and recover the ball after falling. 

New features include cloud accessibility, support for ROS 2 Humble and Windows, improved motion generation extension for robot arms and the ability to add human characters to simulation environments.

NVIDIA used its Isaac Gym RL robotics simulator to train an Allegro Hand, a lightweight, anthropomorphic robotic hand.

The RoboBusiness conference is designed to give technical professionals the information they need to successfully develop, deploy and manage commercial robots.

Conference sessions at RoboBusiness are designed to give technical professionals the information they need to successfully develop, deploy and manage commercial robots.

The US is restricting the sale of the most powerful artificial intelligence processors to China.

With WHIRL, developed at Carnegie Mellon University (CMU), a robot can learn to perform household tasks just watching a person perform them. 

A research team at the robotics department at Google has been developing a new way for robots to learn by watching.