Category Physics

A Technique for more Effective Multipurpose Robots

Four photos show, on top level, a simulation of a robot hand using a spatula, knife, hammer and wrench. The second row shows a real robot hand performing the tasks, and the bottom row shows a human hand performing the tasks.
Caption: Three different data domains — simulation (top), robot tele-operation (middle) and human demos (bottom) — allow a robot to learn to use different tools.
Credits:Image: Courtesy of the researchers

With generative AI models, researchers combined robotics data from different sources to help robots learn better. MIT researchers developed a technique to combine robotics training data across domains, modalities, and tasks using generative AI models. They create a combined strategy from several different datasets that enables a robot to learn to perform new tasks in unseen environments.

Let’s say you want to train a robot so it understands how to use tools and can then quickly learn to make repairs around your house with a hammer, wrench, and screwdriver...

Read More

Controlling Ion Transport for a Blue Energy future: Research highlights the potential of Nanopore Membranes

Controlling ion transport for a blue energy future
Schematic illustration depicting gate voltage control of ion selectivity in a nanopore. Credit: Makusu Tsutsui

Blue energy has the potential to provide a sustainable alternative to fossil fuels. In simple terms, it involves harnessing the energy produced when the ions in a salt solution move from high to low concentrations.

A team including researchers from Osaka University has probed the effect of voltage on the passage of ions through a nanopore membrane to demonstrate greater control of the process.

In a study recently published in ACS Nano the researchers looked at tailoring the flow of ions through the array of nanopores that make up their membrane, and how this control could make applying the technology on a large scale a reality.

If the membranes are made from a charged...

Read More

Bio-Inspired Cameras and AI Help Drivers Detect Pedestrians and Obstacles Faster

Artificial intelligence (AI) combined with a novel bio-inspired camera achieves 100times faster detection of pedestrians and obstacles than current automotive cameras. This important step for computer vision and AI and can greatly improve the safety of automotive systems and self-driving cars.

It’s every driver’s nightmare: a pedestrian stepping out in front of the car seemingly out of nowhere, leaving only a fraction of a second to brake or steer the wheel and avoid the worst. Some cars now have camera systems that can alert the driver or activate emergency braking. But these systems are not yet fast or reliable enough, and they will need to improve dramatically if they are to be used in autonomous vehicles where there is no human behind the wheel.

Quicker detection using less ...

Read More

Transparent Metamaterial for Energy-efficient Regulation in Building can Clean Itself like a Lotus Leaf

Innovative material for sustainable building
Cooling, light-transmissive, and glare-free: the new material combines several unique properties. Credit: Gan Huang, KIT

Researchers at the Karlsruhe Institute of Technology (KIT) introduce a polymer-based material with unique properties in the journal Nature Communications. This material allows sunlight to enter, maintains a more comfortable indoor climate without additional energy, and cleans itself like a lotus leaf. The new development could replace glass components in walls and roofs in the future. The research team has successfully tested the material in outdoor tests on the KIT campus.

Maximizing natural light in buildings is popular and can save on energy costs. However, traditional glass roofs and walls also present problems such as glare, lack of privacy, and overheating...

Read More