Left alone, the toy robots will skitter mindlessly across a tabletop and amuse small children and cats. But when engineers at Princeton paired the small toys with a flexible tether, the bots developed remarkable abilities. They explored enclosed spaces, solved mazes with ease, and even rounded up loose objects into patterns.
A video game in which participants herded virtual cattle has furthered our understanding of how humans make decisions on movement and navigation, and it could help us not only interact more effectively with artificial intelligence, but even improve the way robots move in the future.
Since the industrial revolution, manufacturing processes have continuously evolved in alignment with technological advances. Recent innovations, particularly in the field of robotics, 3D printing and machine learning, could soon facilitate further change, potentially establishing a new generation for industry standards.
Researchers at the National University of Singapore have developed a new robot inspired by one of the most intelligent aquatic animals on Earth: the octopus. This robot, presented in a paper published on the arXiv pre-print server, could be used both to complete real-world tasks underwater and to study the bio-mechanical underpinnings of octopus swimming.
Robots can navigate efficiently through crowds of people by cleverly alternating between independent and cooperative behavior, and in such a way that they disturb the people around them as little as possible. This is the result of a study by TU Professor Roderich Groß posted to the arXiv preprint server.
A pitiful sound from tinny speakers, sad virtual eyes, trembling robot arms: It doesn't take much to feel sorry for a robot. This is the conclusion of a study by Marieke Wieringa, who will be defending her Ph.D. thesis at Radboud University on 5 November. But she warns that our human compassion could also be exploited; just wait until companies find a revenue model for emotional manipulation by robots.
A research team led by Associate Professor Li Mujun, Professor Zhang Shiwu, and Professor Hu Bing from the University of Science and Technology of China (USTC) of the Chinese Academy of Sciences (CAS) has developed porous magnetic soft grippers (PMSGs) that can gently and quickly grasp delicate living things. They can handle a variety of objects, from thin wires to fragile organisms and have potential applications in biomedicine and scientific research. The findings are published in Advanced Materials.
Jeremy Ford hates wasting water. As a mist of rain sprinkled the fields around him in Homestead, Florida, Ford bemoaned how expensive it had been running a fossil fuel-powered irrigation system on his five-acre farm—and how bad it was for the planet.
In the classic cartoon "The Jetsons," Rosie the robotic maid seamlessly switches from vacuuming the house to cooking dinner to taking out the trash. But in real life, training a general-purpose robot remains a major challenge.
KAIST researchers have unveiled a new wearable robot developed for completely paralyzed persons that can walk to them so that the user can wear it right out of their wheelchairs without the help from others. Also, it was announced that Professor Kyoungchul Kong's team from KAIST will be participating in the wearable robot category of the 3rd Cybathlon, which is being held four years after the team's gold medal win in 2020.
Science laboratories across disciplines—chemistry, biochemistry and materials science—are on the verge of a sweeping transformation as robotic automation and AI lead to faster and more precise experiments that unlock breakthroughs in fields like health, energy and electronics.
Imagine sitting in a dark movie theater wondering just how much soda is left in your oversized cup. Rather than prying off the cap and looking, you pick up and shake the cup a bit to hear how much ice is inside rattling around, giving you a decent indication of if you'll need to get a free refill.
For robots, simulation is a great teacher for learning long-horizon (multi-step) tasks—especially compared to how long it takes to collect real-world training data.
In the current AI zeitgeist, sequence models have skyrocketed in popularity for their ability to analyze data and predict what to do next. For instance, you've likely used next-token prediction models like ChatGPT, which anticipate each word (token) in a sequence to form answers to users' queries. There are also full-sequence diffusion models like Sora, which convert words into dazzling, realistic visuals by successively "denoising" an entire video sequence.
In the current AI zeitgeist, sequence models have skyrocketed in popularity for their ability to analyze data and predict what to do next. For instance, you've likely used next-token prediction models like ChatGPT, which anticipate each word (token) in a sequence to form answers to users' queries. There are also full-sequence diffusion models like Sora, which convert words into dazzling, realistic visuals by successively "denoising" an entire video sequence.