Categories
NONE Robotics SCI TECH TECH_Technology

Artificial muscle makes soft robots stronger

Scientists from Harvard University and MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have created artificial muscles that allow soft robots to lift objects that are up to 1,000 times their own weight, a new study published in the Proceedings of the National Academy of Sciences reports.

Soft robotics has made large strides over the past decade. However, while recent advancements have enabled the machines to bend and flex in new ways, the softer materials typically come with reduced strength.

The new origami-inspired muscles in the study get around that obstacle and could one day lead to much more efficient machines.

“We were very surprised by how strong the actuators [aka, “muscles”] were,” said study co-author Daniela Rus, the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science at MIT, according to Phys.org. “We expected they’d have a higher maximum functional weight than ordinary soft robots, but we didn’t expect a thousand-fold increase. It’s like giving these robots superpowers.”

Making muscle-like actuators is one of the largest challenges in engineering. Now that it has been overcome, scientists can potentially build nearly any robot for almost any task.

Each artificial muscle consists of an inner “skeleton” made from materials like metal coil or a sheet of folded plastic surrounded by air or fluid and sealed inside a plastic or textile bag. A vacuum inside the bag causes the muscles to move by forcing the “skin” to collapse onto the skeleton. That tension drives the motion, and allows the device to work without any other external human input. 

In the study, the team created dozens of different muscles with materials ranging from metal springs to packing foam to sheets of plastic. They then experimented with different skeleton shapes to create muscles that can contract down to 10 percent of their original size, lift a flower off the ground, and twist into a coil.

Those experiments showed the muscles can move in many ways, and are able to operate with a high amount of resilience. Not only that, but the technology can generate roughly six times more force per unit area than mammalian skeletal muscle, and is both lightweight and easy to make. A single muscle can be constructed within ten minutes using materials that cost less than $1.

Another important property is that the actuators are highly scalable, meaning they can be constructed at different sizes. That is important because it greatly increases their potential applications. The team believes they could one day be used for a wide variety of tasks, including miniature surgical devices, wearable robotic exoskeletons, transformable architecture, deep-sea manipulators, and large deployable structures for space exploration.

“The possibilities really are limitless,” added Rus, in a statement. “But the very next thing I would like to build with these muscles is an elephant robot with a trunk that can manipulate the world in ways that are as flexible and powerful as you see in real elephants.”

Categories
Research Robotics SCI TECH_Technology

Eagled-eyed machine learning algorithm outperforms human experts

University of Wisconsin-Madison and Oak Ridge National Laboratory researchers just trained artificial intelligence to consistently and quickly analyze and detect microscopic radiation damage in materials considered for nuclear reactors better than human experts.

“Machine learning has great potential to transform the current, human-involved approach of image analysis in microscopy,” said Wei Li, who participated in the research.

“In the future, I believe images from many instruments will pass through a machine learning algorithm for initial analysis before being considered by humans,” said engineering professor Dane Morgan, Li’s graduate school advisor.

The job in question is crucial for the development of safe nuclear materials and could make the time-consuming process more effective and efficient.

“Human detection and identification is error-prone, inconsistent and inefficient. Perhaps most importantly, it’s not scalable,” Morgan said. “Newer imaging technologies are outstripping human capabilities to analyze the data we can produce.”

After training the machine with 270 images, the neural network, in combination with a cascade object detector machine learning algorithm, was able to identify and classify about 86 percent of dislocation loops in a set of sample pictures. In comparison, human experts only found 80 percent of the defects.

“When we got the final result, everyone was surprised, not only by the accuracy of the approach, but the speed,” said Oak Ridge staff scientist Kevin Field. “We can now detect these loops like humans while doing it in a fraction of the time on a standard home computer.”

“This is just the beginning,” Morgan said. “Machine learning tools will help create a cyber infrastructure that scientists can utilize in ways we are just beginning to understand.”

Categories
HEALTH Research Robotics SCI

Robots might hold key to social success for kids with autism

A new study suggests that autonomous “social” robots can be utilized in home-based therapy to model and encourage behaviors to children with autism spectrum disorder. The behaviors included paying attention and maintaining eye contact.

According to the mother of one boy who took part in the experiment, she observed tangible results in her son, who is typically “a little awkward at times” with people he is not close to.

“From watching the robot and interacting with the robot, that really engaged him, and it made him, I think, connect the dots. His interactions became more consistent,” she said. “His eye contact became more consistent.”

“It really just showed me how bright he is and how quick he is,” she continued. “And it gave us time together, to kind of learn about each other. He’s a lot of fun, and this really brought out really good qualities for him.”

“The robot acted in part as the game moderator, selecting appropriate difficulty levels, posing new challenges [and] advancing the narrative of the games,” said study leader Brian Scassellati, who is from Yale University.

Thomas Frazier, chief science officer for Autism Speaks, believes that the study is an “important advance.”

“The authors are careful to note several of the current limitations, including the relatively restricted context in which the robot is being used,” he said. “But, at the same time, this provides a nice advance toward the ultimate goal of personalized support throughout the day and in various settings.”

“The fact that engagement was so high, performance improved and caregivers felt that the child’s behavior improved is impressive,” he added.

The findings were published in Science Robotics.

Categories
HEALTH HND_Disease Robotics SCI

Cannabis-like drug could help Alzheimer’s patients, study says

A new study suggests that a cannabis-like drug called nabilone might treat agitation in patients with Alzheimer’s disease. The synthetic drug is typically used to control nausea in cancer patients.

Agitation is one of the most disturbing Alzheimer’s symptoms and it can be very tough to manage. In order to control the symptom, many doctors go against medical advice and prescribe antipsychotic drugs and sometimes even physically restrain patients.

“Agitation, aggression, sleep disturbances — all have a significant impact on both their quality of life and their carers’ quality of life,” said Heather Snyder, senior director of medical operations for the Alzheimer’s Association.

“Currently prescribed treatments for agitation in Alzheimer’s do not work in everybody, and when they do work the effect is small and they increase risk of harmful side effects, including increased risk of death,” said Krista Lanctôt of the University of Toronto, who led the research. “As a result, there is an urgent need for safer medication options.”

Lanctôt are her team tested nabilone for six weeks by giving the pill to 39 dementia patients. Afterwards, they gave them a placebo for an additional six weeks.

“Agitation improved significantly in those taking nabilone, compared to placebo,” the Alzheimer’s Association said in a summary of the research, which is being presented at its annual Chicago meeting.

The non-profit association also said that “nabilone also significantly improved overall behavioral symptoms, compared to placebo, as measured by the Neuropsychiatric Inventory” questionnaire.

The findings were published in The American Journal of Geriatric Psychiatry.

Categories
Robotics TECH_Technology

Deep reinforcement learning is the latest trend in AI technology

A new area of AI research, called Deep reinforcement learning (DRL) is making waves within artificial general intelligence (or AGI) circles, writes Sam Charrington for Venture Beat.  This is because DRL mirrors human learning by exploring and receiving feedback from environments.

Supervised machine learning trains models based on “known-correct” answers.  By contrast, researchers implement reinforcement learning by having an “agent” interact with an environment.  Thus, when the agent’s actions produce a desired result, it receives positive feedback.  The promise of DRL has led to a number of startups hoping to capitalize on this technology, writes Charrington.  Pieter Abbeel at the University of California Berkeley has founded Embodied Intelligence, a startup that will combine VR (virtual reality) and DRL and apply it to robotics.  Pit.ai plans to out-trade traditional hedge funds by applying it to algorithms.

Increased interest has also led to startups creating new open source toolkits and environments for training DRL agents.  Charrington lists several of these interfaces, like House3D (a collaboration between UC Berkeley and Facebook AI researchers).  It offers over 45,000 simulated indoor scenes with realistic room and furniture layouts.  The primary task, introduced in the team’s paper, is “concept-driven navigation,” which would train an agent to navigate to rooms within a house.

Charrington cautions that applying DRL to non-trivial problems creates the challenge of constructing a reward function that encourages desired behaviors without the adverse effect of promoting cheating.  However, with all the tools and platforms in development, researchers hope to eventually work through these challenges.

Categories
Robotics TECH TECH_Technology

DJI Mavic 2 drones add multi-sided obstacle sensors

DJI has announced its first consumer drones to feature obstacle-detecting sensors on all sides of their bodies.

The Mavic 2 aircraft use 10 sensors to prevent collisions with objects above and below them, as well as threats in front or behind and to their left or right.

The foldable aircraft also offer improved cameras over their predecessors, including a zoom lens in one and better color via the other.

The Chinese company had originally planned to launch the products in July and its decision to postpone came too late to prevent many of the details being printed in retailer Argos’s catalog.

“It also faces competition from two other models whose rotor blades can also be tucked away when not in use – Yulee’s Mantis Q and Parrot’s Naïf,” BBC reports.

However, GoPro announced earlier this year that it would not sell a successor to its foldable Karma drone, which competed with the original Mavic models.

Shenzhen-based DJI has a 74% share of the global drone market, according to data released by the analysis company Skylogic earlier this month.

Safety concerns play a part in limiting demand, with no-fly zones imposed over many built-up areas and parks.

“Airspace restrictions are the number one complaint that people have about their drones,” commented Skylogic’s chief executive Colin Snow.

“But in the US we are making improvements – the aviation authorities understand the issue.”

The UK is among countries currently considering requiring owners to register their models and pass a test before using them, as well as the possibility of a minimum age.

Categories
Robotics TECH_Technology

Europe aims to compete in the global AI research market with 1.5 billion Euros investment

The European Commission has announced that it is devoting €1.5 billion to AI research funding through 2020 in an effort to catch up to the United States and China. The commission also plans to release ethical guidelines for AI development by the end of the year, reports Tania Rabesandratana for Science.

French President Emmanuel Macron weighed in on the investment, pledging France’s commitment. “We have to be in a position to build, in France and in Europe, an artificial intelligence ecosystem,” he said. “We should have a policy of open data,” he added, and “have to think on the subject from a political and ethical point of view … to come up with a common understanding and rules.” The commission says it will fund basic research, and it intends to help member states set up joint research centers across Europe.

Both the United States and China practice “permissionless innovation,” says Eleonore Pauwels, a Belgian ethics researcher at the United Nations University. Europeans, in contrast, want to develop “AI systems that require smaller data sets to enhance privacy and trust, and are more transparent than their competitors,” she says. Jeffrey Ding studies AI governance at the University of Oxford, and monitors the AI potential of different countries. As he explains it, Europe has strong AI research, but a weak AI industry partly because venture capital funding of startups in the United States and China far exceeds that of Europe.

Twenty-five European countries signed a declaration on April 10th to work together and set AI research funding “as a matter of priority.”

Categories
NWT_Animals Robotics

Jumping spiders could one day lead to better robots

By training a spider to jump, researchers from the University of Manchester managed to gain new insights into the mechanisms behind arachnid movement.

In the study, the team found that the arachnid — nicknamed Kim — could jump six times her body length from a standing start. In contrast, humans can only jump 1.5 body lengths.

Kim is a regal jumping spider (Phidippus regius), a species known for its ability to pounce on prey. To study the impressive jumping, researchers filmed the spider with high-tech cameras and then used 3D CT scans to build a model for both her legs and body structure.

That revealed Kim used different jumping strategies at different times. For instance, sometimes she would use a faster jump with a lower trajectory for increased accuracy, and sometimes she would tend towards more energy efficient ones for longer distance.

“She will jump at the optimal angle, which means that she can understand the challenge that she is presented with,” explained lead author Mostafa Nabawy, a researcher at the University of Manchester, according to BBC News“And then she can time her jumping performance at take-off to execute a jump that is optimal in terms of energy demand.”

The data collected from the study gives new insight into the forces behind a spider’s jump, as well as how such forces are generated. That could then shed light on certain biomechanics and allow scientists to use them in other areas of research.

As Kim only used muscle power to jump, the impressive leaping ability may then one day lead to a new generation of more efficient robots or other such machines.

“We aim to use this improved understanding of spiders to imagine a new class of agile micro-robots that are currently unthinkable using today’s engineering technologies,” said Nawaby, according to Gizmodo.

 

The research is published in the journal Scientific Reports.

Categories
NWT_Biology Robotics TECH TECH_Technology

New self-healing skin could lead to advanced robotics

A team of researchers at the University of Colorado, Boulder have created an electronic skin that can heal itself over time, according to new research in Science Advances

The new technology is a thin film equipped with a series of advanced sensors that help it detect and interact with the environment. Not only can it detect pressure, but it has the ability to measure temperature, humidity, and air flow as well. In that way, it acts like human skin.

To develop the material, researchers used a covalently bonded dynamic network polymer known as polyimine. They laced the substance with silver nanoparticles that allow it to heal itself when cut or damaged. It does that by recreating chemical bonds between the two split pieces, giving them a way to comeback together.

Another benefit of the material is that it is easily recyclable. If it becomes damaged beyond repair, scientists can simply soak it in a special solution that breaks it down for future use.

“To recycle the skin, the device is soaked into recycling solution, making the polymers degrade into oligomers, polymers with polymerization degree usually below 10, and monomers, small molecules that can be joined together into polymers, both of which are soluble in ethanol,” said study co-author Xiao Jianliang, assistant professor from University of Colorado Boulder, according to Perfscience. “The silver nanoparticles sink to the bottom of the solution.

The new skin is a great advancement in electronic organs, but its biggest application could be robotics. There is still a lot of undiscovered territory in the field, and being able to create skin that is sensitive to touch could help create much more gentle and subtle machines.

“Let’s say you wanted a robot to take care of a baby,” explained study co-author Wei Zhang, a researcher at the University of Colorado, Boulder. “In that case you would integrate e-skin on the robot fingers that can feel the pressure of the baby. The idea is to try and mimic biological skin with e-skin that has desired functions.”

Categories
HEALTH Robotics TECH_Technology

Will robots replace surgeons?

Robots already are performing surgery on humans. This technology will most likely become the wave of the future and a preferred option for informed patients.

Earlier this month Dr. Greg Shaw, a consultant urologist/surgeon removed a cancerous prostate gland from a 59-year-old man at University College Hospital in London, England. He used four dextrous metal arms of an American-made machine with his head under the black hood of a 3D monitor.

Shaw has performed 500 such surgeries. He says they are particularly useful for pelvic operations. The process involves:

  • Docking the cart onto the human
  • Inserting three surgical tools and a video camera in small incisions in the patient’s abdomen
  • Clipping and cauterizing

This hands-off operation is part of a clinical trial that aims to preserve the architecture of microscopic nerves around the prostate. It also preserves a patient’s sexual function.

Shaw uses a Da Vinci robot for the surgery. The American firm Intuitive Surgical makes the cutting edge robot that enables the UCH team to do 600 prostate operations a year.

Other advantages of this technology include 1. it lends itself to accelerated and effective training because it retains a 3D film of all operations. 2. It enables virtual-reality suite for medical students to use.

According to Lord Darzi, the surgeon and former minister in Gordon Brown’s administration, many functions traditionally performed by doctors and nurses could be replaced by robots. Patients are becoming more aware of these alternatives to conventional surgical procedures.