Sunday, March 9, 2025
Home Blog Page 1457

Magnetically driven soft robot achieves high-speed jumping

0
Magnetically driven soft robot achieves high-speed jumping


A magnetically driven ultrafast bistable soft jumping robot
Image outlining the proposed bistable soft jumper and its fabrication. Credit: Tang et al. (Science Robotics, 2024).

Many animal species, ranging from insects to amphibians and fish, use jumping as a means of moving within their surrounding environment. Jumping can be very advantageous for these animals, for instance, allowing them to reach higher branches of trees, swiftly escape from predators or move faster across long distances.

Many roboticists have been trying to develop robots that can replicate the jumping locomotion styles observed in animals, as these robots could have interesting real-world applications. By jumping, robots could move faster on complex terrains and gain access to surfaces or environments that they might otherwise be unable to reach.

The jumping robots introduced in recent years rely on various actuating methods, ranging from dielectric elastomers to liquid crystal elastomers and soft actuators. While some of these robots achieved promising results, most of them were found to lag behind living organisms that are highly skilled jumpers, both in terms of how high and how fast they can jump.

Researchers at Zhejiang University in China recently developed a new ultrafast, magnetically driven and bistable soft jumper that demonstrated advanced jumping capabilities. This jumper, presented in a paper published in Science Robotics, was found to achieve different jumping locomotion styles, jumping higher and faster than comparable robotic systems introduced in the past.






Credit: Daofan Tang

Soft jumpers, such as the system developed by these researchers, are based on elastic and deformable materials that often have a greater resistance to impact, preventing damage to the robot while jumping. Yet many existing jumpers based on soft materials were found to be limited in terms of the speed with which they respond to stimuli and takeoff from the ground.

“We report a magnetic-driven, ultrafast bistable soft jumper that exhibits good jumping capability (jumping more than 108 body heights with a takeoff velocity of more than 2 meters per second) and fast response time (less than 15 milliseconds) compared with previous soft jumping robots,” wrote Daofan Tang, Chengqian Zhang and their colleagues in their paper. “The snap-through transitions between bistable states form a repeatable loop that harnesses the ultrafast release of stored elastic energy.”

The researchers created prototypes of their jumper that varied in size and found that smaller jumpers were more affected by air resistance; thus they could not jump as high as bigger jumpers. Nonetheless, the takeoff velocities of the jumpers remained similar, irrespective of their size.

A magnetically driven ultrafast bistable soft jumping robot
Schematic illustration of bistable soft jumper. Credit: Daofan Tang

Notably, the jumper designed by this research team can perform two different types of locomotion, namely jumping and hopping. The researchers carried out tests in a real-world environment to demonstrate the advantages of these locomotion modes.

“These modes are controlled by adjusting the duration and strength of the magnetic field, which endows the bistable soft jumper with robust locomotion capabilities,” wrote Tang, Zhang and their colleagues. “In addition, it is capable of jumping omnidirectionally with tunable heights and distances. To demonstrate its capability in complex environments, a realistic pipeline with amphibious terrain was established.”

The researchers tested their jumper in a simple locomotion task that entailed hopping through a narrow tube, jumping through a U-shaped pipeline, and jumping from underwater to above the water level. This task was designed to simulate a scenario in which the robot could be used to clean water inside a pipeline.

In this initial experiment, the mechanically driven jumper was found to perform remarkably well. In the future, its underlying design could inspire the development of other flexible robotic systems for a wide range of real-world applications.

More information:
Daofan Tang et al, Bistable soft jumper capable of fast response and high takeoff velocity, Science Robotics (2024). DOI: 10.1126/scirobotics.adm8484.

© 2024 Science X Network

Citation:
Magnetically driven soft robot achieves high-speed jumping (2024, September 16)
retrieved 16 September 2024
from https://techxplore.com/news/2024-09-magnetically-driven-soft-robot-high.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.





Source link

New algorithm rights wrongs of precipitation-type classification over Tibetan Plateau

0
New algorithm rights wrongs of precipitation-type classification over Tibetan Plateau


Righting wrongs of precipitation-type classification over Tibetan plateau with new algorithm applied to satellite radar data
The cover of the Advances in Atmospheric Sciences. Credit: USTC, Yunfei Fu.

Like many natural phenomena, precipitation can be both a blessing and a scourge to human life. On the one hand, it supplies our rivers and fields with water; on the other hand, it can cause floods, landslides, and other natural disasters. Either way, understanding and predicting the different types of precipitation is essential.

Gathering and analyzing precipitation data is key, but there are some places on Earth where this is tricky. One such place is the Tibetan Plateau, whose unique and challenging physical environment makes observing precipitation, either on the ground or from above via satellites, a problematic issue.

In terms of satellite-borne precipitation radar, the high altitude of the Tibetan Plateau leads to cases of mistaken identity when it comes to the specific types of precipitation. Most notably, because the height of the terrain on the plateau is close to the height of the freezing level in the atmosphere over non-plateau areas, weak convective precipitation can be misclassified as stratiform precipitation.

Recognizing this, in a recent study published in Advances in Atmospheric Sciences, Prof. Fu Yunfei from the University of Science and Technology of China (USTC), along with colleagues from the China Meteorological Administration, first analyzed in detail the problems with the existing precipitation-type identification algorithm for satellite-borne precipitation radar, wherein they uncovered the reasons why the algorithm can fail in its identification of precipitation types over the Tibetan Plateau in summer, and then developed and tested a new one.

“It often seems that there is a desire to standardize everything in the natural world, while ignoring the obvious intricacies and diversities that exist at various scales,” explains Fu. “In our scientific field, meteorology, this also holds true; and in terms of the specific problem that we set out to solve in this study, the thresholds employed to identify types of precipitation are a specific example.”

Traditionally, based on observations in non-plateau regions, the approach to classifying precipitation is a simplistic one, ending with a somewhat binary result whereby precipitation is identified either as convective or stratiform.

With the new algorithm developed by Fu’s team, parameters such as maximum reflectivity factor, background maximum reflectivity factor, and echo top height are considered in greater depth to yield a more granular classification (types include “strong convective,” “weak convective,” “weak,” and “other”) that provides more useful information with far fewer identification errors.

Ultimately, the work of Fu and his team can be carried forward and adopted by the weather forecasting and modeling community to better predict the occurrence of different types of precipitation events in parts of the world where existing, standardized methods are less applicable. More specifically, as highlighted by the context in this particular paper, there are major benefits to be felt by communities living in mountainous regions like the Tibetan Plateau.

“It should be said, however,” ends Fu, “that there is still much to be done. In particular, we need to confirm the existence of stratiform precipitation over the Tibetan Plateau in summer. For a variety of reasons, this is still difficult to detect using satellite-borne precipitation radar measurements. This will be the next step in our research.”

The paper is featured on the cover of issue Advances in Atmospheric Sciences. The cover photo was taken by Fu in the Lianbao-Yeze Nature Reserve located at the junction of Xichuan, Gansu and Qinghai provinces, with an average elevation of more than 4,000 meters above sea level. It shows convective clouds over the lake in the foreground, as well as over the far end of the mountain in the center.

More information:
Yunfei Fu et al, A New Algorithm of Rain Type Classification for GPM Dual-Frequency Precipitation Radar in Summer Tibetan Plateau, Advances in Atmospheric Sciences (2024). DOI: 10.1007/s00376-024-3384-7

Citation:
New algorithm rights wrongs of precipitation-type classification over Tibetan Plateau (2024, September 16)
retrieved 16 September 2024
from https://phys.org/news/2024-09-algorithm-rights-wrongs-precipitation-classification.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.





Source link

The silent conversations of plants

0
The silent conversations of plants


plants
Credit: CC0 Public Domain

This morning, my six-year-old came into our bedroom and started reading a story from a book. She followed each word on the page, slowly forming full sentences. Sometimes she stumbled and asked for help with some “funny words,” but by the end of the book, she had told us a story about a bear in the snow.

Verbal communication is one of the many reasons humans have become so successful as a species. From warning each other of danger to communicating complex information, our ability to speak has been crucial.

But it’s not just humans and other animals who have developed sophisticated communication. A lot of people think of plants as passive, but they have their own way of interacting with each other. The idea has been around for a while, even inspiring Hollywood movies like “Avatar.”

However, recent science is showing plant communication systems may be more complex than we imagined.

These communication networks are sensitive and in balance. Imagine how disrupted our world would be if global network systems suddenly broke down. The recent CrowdStrike IT outages are just one example of how delicate these systems are and how important communication is—and that’s the case for plants too.

To grasp how organisms who can’t speak pass on information to each other, it’s important to understand that humans also have a non-verbal communication system. This includes our senses of sight, smell, hearing, taste and touch.

For example, natural gas companies add a chemical called mercaptan to natural gas, giving it that distinctive “rotten egg” smell to warn us of leaks. Think also of how we have developed sign language, while many people are skilled lip readers.

In addition to these senses, we also have equilibrioception (the ability to maintain balance and body posture), proprioception (the sense of the relative position and strength of our body parts), thermoception (sense of temperature changes), and nociception (ability to sense pain). All these abilities have enabled humans to become highly sophisticated in communication and engagement with the natural world.

Other species, particularly plants, use their senses to spread information in their own way.

What are the neighbors up to?

Most of us are familiar with the smell of freshly cut grass. The volatiles, or chemical substances, released by the grass plants, which we associate with that smell, are one way they communicate to other nearby plants that a predator—or in this case a lawnmower—is present, prompting an adjustment in plant defenses. Rather than using auditory cues, plants use chemical-induced communication. However, plant communication doesn’t end with volatiles.

Recently, scientists discovered just how well-connected plants are and how efficiently they can send messages to their peers via their roots, electrical signals, a network of underground fungi and soil microbes. The nosy plant neighborhood watch was discovered.

For example, electrophysiology is a relatively new scientific discipline that studies how electrical signals in and between plants are communicated and interpreted. With major advances in technology and artificial intelligence (AI), we have seen significant accelerated growth in this area of research in the past few years.

Scientists could be on the verge of remarkable discoveries, with recent advances integrating electrical signal communication within and between plants into modern greenhouses to monitor and control crop watering or detect nutritional deficiencies.

Scientists achieve this by inserting small electrical probes, similar to acupuncture needles, to test how changes in electrical signals relate to plant performance, such as transporting water and nutrients, and converting light into important sugars.

Researchers have even influenced plant behavior by sending electrical signals from mobile phones, making them perform basic responses like opening or closing leaves in a Venus flytrap.

Soon we may be able to fully translate the language of our crops.

A great deal of plant-to-plant communication happens below ground, facilitated by large fungal networks known as the “wood-wide web.” This network of fungi connects trees and plants underground, allowing them to share resources like water, nutrients and information. Through this system, older trees can help younger ones grow, and trees can warn each other about dangers such as pests.

It’s like an underground internet for trees and plants, helping them support and communicate with each other. The network is extensive, with over 80% of plants believed to be connected, making it one of the oldest communication systems in the world.

Just as the internet enables us to connect, share ideas, knowledge and information that can influence decision-making, the “wood-wide web” allows plants to use symbiotic fungi to prepare for environmental changes.

However, disturbing the soil through chemicals, deforestation or climate change can disrupt the communication nodes through affecting water and nutrient cycles in these networks, making plants less informed and connected. Not much research has yet been carried out into the effects of disrupting these networks.

But we know that plants’ responsive behavior, such as defense responses and gene regulation, can be altered by their fungal network if they are connected to one.

So this communication-disconnect might make them more vulnerable, making it more difficult to protect and restore ecosystems around the world. There is still a lot scientists have to learn about these highly complex networks

We know it’s important to help children learn to read so that they can navigate the world around them. It is just as important as to ensure we do not disconnect plant communication. After all, we depend upon plants for our well-being and survival.

Provided by
The Conversation


This article is republished from The Conversation under a Creative Commons license. Read the original article.The Conversation

Citation:
The silent conversations of plants (2024, September 16)
retrieved 16 September 2024
from https://phys.org/news/2024-09-silent-conversations.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.





Source link

Solving long-standing challenge in semiconductor manufacturing—a refined algorithm for detecting wafer defects

0
Solving long-standing challenge in semiconductor manufacturing—a refined algorithm for detecting wafer defects


Minting wafer thin defect detection
Illustration of SPD-Conv structure with partition factor P = 2. Credit: International Journal of Information and Communication Technology (2024). DOI: 10.1504/IJICT.2024.141433

Research published in the International Journal of Information and Communication Technology may soon help solve a long-standing challenge in semiconductor manufacture: the accurate detection of surface defects on silicon wafers. Crystalline silicon is the critical material used in the production of integrated circuits and in order to provide the computing power for everyday electronics and advanced automotive systems needs to be as pristine as possible prior to printing of the microscopic features of the circuit on the silicon surface.

Of course, no manufacturing technology is perfect and the intricate process of fabricating semiconductor chips inevitably leads to some defects on the silicon wafers. This reduces the number of working chips in a batch and leads to a small, but significant proportion of the production line output failing.

The usual way to spot defects on silicon wafers has been done manually, with human operators examining each wafer by eye. This is both time-consuming and error-prone due to the fine attention to detail required. As wafer production has ramped up globally to meet demand and the defects themselves have become harder to detect by eye, the limitations of this approach have become more apparent.

Chen Tang, Lijie Yin and Yongchao Xie of the Hunan Railway Professional Technology College in Zhuzhou, Hunan Province, China explain that automated detection systems have emerged as a possible solution. These too present efficiency and accuracy issues in large-scale production environments. As such, the team has turned to deep learning, particularly convolutional neural networks (CNNs), to improve wafer defect detection.

The researchers explain that CNNs have demonstrated significant potential in image recognition. They have now demonstrated that this can be used to identify minute irregularities on the surface of a silicon wafer. The “You Only Look Once” series of object detection algorithms is well known for being able to balances accuracy against detection speed.

The Hunan team has taken the YOLOv7 algorithm a step further to address the specific problems faced in wafer defect detection. The main innovation in the work lies in using SPD-Conv, a specialized convolutional operation to enhance the ability of the algorithm to extract fine details from images of silicon wafers. Additionally, the researchers incorporated a Convolutional Block Attention Module (CBAM) into the model to sharpen the system’s focus on smaller defects that are often missed in manual inspection or by other algorithms.

When tested on the standard dataset (WM-811k) for assessing wafer defect detection systems, the team’s refined YOLOv7 algorithm achieved a mean average precision of 92.5% and had a recall rate of 94.1%. It did this quickly, at a rate of 136 images per second, which is faster than earlier systems.

More information:
Chen Tang et al, Wafer surface defect detection with enhanced YOLOv7, International Journal of Information and Communication Technology (2024). DOI: 10.1504/IJICT.2024.141433

Citation:
Solving long-standing challenge in semiconductor manufacturing—a refined algorithm for detecting wafer defects (2024, September 16)
retrieved 16 September 2024
from https://techxplore.com/news/2024-09-semiconductor-refined-algorithm-wafer-defects.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.





Source link

TikTok battles US ban threat in court

0
TikTok battles US ban threat in court


Social media app TikTok has come under scrutiny from the US government
Social media app TikTok has come under scrutiny from the US government.

TikTok faced pushback in a federal court on Monday in its efforts to stop a law that requires the app to divest from its Chinese ownership or face a ban in the United States.

A three-judge panel of the US Court of Appeals in Washington heard arguments from TikTok, its owner ByteDance, and a group of users claiming that the ban violates free speech and is unconstitutional.

The US government alleges TikTok allows Beijing to collect data and spy on users. It also says TikTok is a conduit to spread propaganda. China and the company strongly deny these claims.

TikTok has until January to find a buyer or face the ban, which would likely provoke a strong response from the Chinese government and further strain US-China relations.

It would also upend the social media business and rile many of the app’s 170 million US users.

ByteDance, TikTok’s parent company, has stated it has no plans to sell TikTok, leaving the app’s legal appeal—focused on US guarantees for free speech—as its only option for survival.

“The law before this court is unprecedented. Its effect would be staggering,” said Andrew Pincus, the lawyer arguing on behalf of the wildly popular video-sharing app.

“For the first time in history, Congress has expressly targeted a specific US speaker (i.e., TikTok U.S.),” he added.

In their questions, the judges challenged this argument, comparing it to earlier cases in US jurisprudence.

This included a case from the 1980s where closing the Palestine Information Office in Washington DC was deemed legal because it was backed by the PLO, an organization officially designated as a terrorist group.

TikTok’s lawyer countered: “Mere foreign ownership can’t possibly be a justification, because it would turn the First Amendment (protecting free speech) on its head.”

He added that seeing foreign ownership alone as criteria for forced divestiture “would be a pretty shocking change here,” citing other foreign-owned media companies such as Politico, Al Jazeera, and the BBC.

The lawyer also questioned why the US law did not target e-commerce sites with similar Chinese ownership.

Pincus said that if you followed the US government’s logic, which he disagreed with, “certainly those sites could well be susceptible to (China’s) action, but they’ve been excluded by Congress (in the law).”

‘Important questions’

The judges grilled the US government on whether TikTok U.S., a US-based company, should be denied its free speech rights.

The US government lawyer, Daniel Tenny, insisted that the content being targeted was a recommendations algorithm based at ByteDance in China, not anything created by US users, and that it was therefore out of reach of free speech considerations.

“There’s really no dispute here that the recommendation engine is maintained, developed, and written by ByteDance, rather than TikTok US, and that is what’s being targeted,” Tenny argued.

The trio of judges will decide the case in the coming weeks or months, but regardless of their decision, the case is likely to reach the US Supreme Court, experts said.

“After listening to the oral arguments, I am more convinced that this case will end up in the Supreme Court,” said Sarah Kreps, director of Cornell’s Tech Policy Institute.

“Overall, the judges sounded more skeptical of the TikTok case but also raised important questions about the First Amendment, foreign influence and standards of scrutiny that I do not think were clearly resolved with today’s exchanges,” she added.

The fate of Americans’ access to TikTok has become a prominent issue in the country’s political debates, with Republican presidential candidate Donald Trump opposing a ban.

Democratic President Joe Biden, whose vice president Kamala Harris is running against Trump, signed the law that gives TikTok until January to shed its Chinese ownership or be expelled from the US market.

© 2024 AFP

Citation:
TikTok battles US ban threat in court (2024, September 16)
retrieved 16 September 2024
from https://techxplore.com/news/2024-09-tiktok-threat-court.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.





Source link