Disney develops 3D printed 2-legged robot!

http://www.3ders.org/articles/20150527-disney-develops-2-legged-3d-printed-robot-that-walks-like-an-animated-character.html

Disney develops 3D printed 2-legged robot that walks like an animated character

There are just a few companies in the world that need no introduction, and Disney is one of them. After all, who didn’t grow up watching Disney classics? But did you know that Disney does more than shoot box office hits, record terrible catchy songs and avoid theme park-related lawsuits? They also have an active Research Department charged with creating actual, rather than digital, creations which can be used for throughout the Disney imperium. And the department’s latest achievement is impressive: recreating the walking movements of animated characters in bi-pedal robots, which they have done using 3D printing technology.

As three scientists attached to the department in Pittsburgh – Seungmoon Song, Joohyung Kim and Katsu Yamane – explain, they set out to develop robotics that can be used to make Disney’s theme parks and toys more realistic and magical. After all, fit young heros from Disney’s movies and TV shows don’t exactly perform well when moving as stiffly as paraplegic grandmothers. ‘Creating robots that embody animation characters in the real world is highly demanded in the entertainment industry because such robots would allow people to physically interact with characters that they have only seen in films or TV. To give a feeling of life to those robots, it is important to mimic not only the appearance but also the motion styles of the characters,’ they write.

But this isn’t easy. As they write in an article entitled ‘Development of a Bipedal Robot that Walks Like an Animation Character’, the field of robotics struggles to capture life-like movement. ‘The main challenge of this project comes from the fact that the original animation character and its motions are not designed considering physical constraints,’ they write. And of course trying to tackle quirky and fast animated characters is even more difficult, as they movements are not typically designed to be physical correct. ‘[But in recent years] animation characters have evolved to be more realistic. Using computer graphic techniques, we can design 3D characters, and generate more natural and physically plausible motions with them.’

And you might be surprised to learn that their solution is somewhat similar to what you and I would do for a project: just 3D print it and add some servo motors. Of course it isn’t quite so simple, but to capture the exaggerated gait and movement of animated characters they first 3D printed leg components to match the structure of their potato-like character, which you can see in the clip below. ‘We start from animation data of a character walking. We develop a bipedal robot which corresponds to lower part of the character following its kinematic structure. The links are 3D printed and the joints are actuated by servo motors,’ they explain. All these parts were 3D printed using Stratasys’ Object 260 Connect 3D printer in RGD525 material.

Of course these need to be very specifically angled and positioned to ensure that 3D movement can be recreated. And Trajectory optimization software does most of the rest. ‘Using trajectory optimization, we generate an open-loop walking trajectory that mimics the character’s walking motion by modifying the motion such that the Zero Moment Point stays in the contact convex hull,’ they write. Now this process is more difficult than it sounds, but for a full description of data extraction and installing the mechanics you’ll have to dive into the full scientific article here.

But the results are obvious, though not perfect. The robot can definitely walk well, but doesn’t reproduce the digital models perfectly and has a tendency to wobble. ‘When we play back the optimized trajectory, the robot wobbles forward. It is because the robot does not produce the motion perfectly. For example, the stance leg flexes more and scuffs the swing foot at the beginning and end of the swing phase. This causes the swing foot to push against the ground and the stance foot to slip, which results in unstable walking,’ the scientists write.

One solution for this is slowing down the process. ‘We observed that the robot slips less as we play back the optimized motion slower, and the resulting walking looks closer to the optimized walking,’ they write, but conclude that the system just isn’t working optimal for now. While there are few options for more progress – including investigating structural materials and replacing 3D printed parts – it looks like we’ll have to wait a few years before running into mechanically-sound walking Disney characters at Disney world.

3ders.org

by Alec | May 27, 2015

http://www.3ders.org/articles/20150527-disney-develops-2-legged-3d-printed-robot-that-walks-like-an-animated-character.html

3D printed prosthetics and their restrictions

An Eye-Opening Article about 3D Printed Prosthetics & Their Restrictions

http://www.thedailybeast.com/articles/2015/03/11/the-reality-of-3d-printed-robo-hands.html

London, UK. 7th November 2013. a 3D printed prosthetic arm by the University of Nottingham is on display at the 3D Printshow at the Business Design Centre in London.
The Show brings together the biggest names in 3D printing technology alongside the most creative, exciting and innovative individuals using additive processes today. © Piero Cruciatti/Alamy Live News

3D printed prosthetics seems like a miracle solution to a costly problem, but some of the claims of safety and cost are exaggerated.

Jack Reidy just turned 10 last month. He’s an athletic kid—in the winter he plays hockey a few times a week, and in the summer he pitches on his baseball team. His style is a little unorthodox though. When he’s pitching, after he releases the ball, he switches his glove onto his throwing hand. And on the rink, he holds the stick against his body on his left side, rather than in his hand. That’s because Jack was born with a partial left hand, with a palm but without fully formed fingers.

Despite that, Jack’s father, James, said that his son never really asked for a prosthetic. “We never brought it up, we’ve just treated him as any normal kid.”

It wasn’t until last year, when Jack saw a picture of a 3D printed hand in Parademagazine, that he even considered it. “His first thought was holding a baseball bat, and that was Jack’s first time in showing any interest in any type of prosthetics.”

The Reidys had been working with a prosthetist named Jeff Erenstone to help develop a special hockey glove, and it turned out that Erenstone was also involved in a group of volunteers who print plastic hands.

Eventually Jack was matched with a volunteer in Michigan named Bruce Chaput, who offered to print him a hand (a model called the Raptor). James remembers it all being quite foreign to him. “He sent us a picture of the printer. It looked funky; I thought it was some old age kind of thing. Obviously it’s not.”

Chaput and Jack spent a lot of time talking about what colors he wanted his hand to be. “We spent way more time talking about the colors and the hand and whatnot than actually printing it,” Chaput said, laughing. Jack picked orange and black, the colors of the high school where his dad coaches hockey.

On Christmas Day, the hand arrived at Erenstone’s office. But it wasn’t exactly a Christmas miracle. When Erenstone opened the box, he immediately noticed a long crack in the hand. And when he picked it up things got worse. “It literally crumbled in my hand,” he said.

Recently there’s been a lot of hype surrounding the promise of 3D printed limbs. Everywhere from The New York Times to Popular Science to the Today Show has run stories on people all over the world printing hands. The narrative goes like this: Prosthetic hands are really expensive—a recent Uproxx documentary about 3D printed hands claimed that the average prosthetic on the market costs $60,000—while the 3D printed version cost far less, and can be fixed and replaced with a simple push of a button on a printer. Welcome to the future, the world in which the everyman can print his own arm, breaking free from the chains of debt-by-prosthetic.

But that’s not exactly a true story.

Last month, the American Orthotics and Prosthetics Association (AOPA) released a statement clarifying a few key points. The average upper extremity prosthesis does not cost anywhere near $40,000 to $80,000, as many of these accounts claim. It actually costs something like $1,500 to $8,000. The AOPA statement also pointed out that in many cases, the people printing hands are operating illegally. There are 15 states in which providing a prosthetic or orthotic device is illegal without a license. Prosthetists are trained medical professionals, with licenses that take years of education and apprenticeships. The people printing these arms have none of that, which can, in theory, become dangerous. These arms and hands they’re printing aren’t FDA tested, break easily, and should never be used to replace a prosthetic arm.

Of course, the reality of 3D printed prosthetics is somewhere in between the media hype and the concerns of prosthetists with an industry to protect. Prosthetics made by 3D printers can certainly help some people, especially children who are embarrassed of their missing limb. But it’s also important to remember that these are, for the most part, hands made out of thin layers of plastic, printed by volunteer hobbyists with no training.

Some organizations understand that. The group that Erenstone hooked the Reidys up with is one of them, called e-NABLE. e-NABLe is a community-based group that connects amputees with hobbyists who have 3D printers, and is a good example of an organization that understand the limits of their technology.

“We don’t even call these things prosthetics,” Jon Schull, the co-founder of e-NABLE, told me. Schull said they have turned away amputees asking for hands for tasks that they’re not capable of standing up to. “We had someone who used to ride a motorcycle, who wanted hands so he could ride his motorcycle again. He had big hopes for what this could do that we weren’t comfortable with. He was going to use it to operate heavy machinery that could injure himself and others.”

Despite that, Schull said that the group’s relationship with prosthetists is shaky. “Some of them are concerned that we’re undercutting their industry. Some of them understand that we’re opening up a new market.”

The AOPA statement came out of frustration from prosthetists that some 3D printing groups were promoting their work using inaccurate numbers. But Tom Fise, the executive director of AOPA, said that he’s not trying to discourage groups like e-NABLE from doing the work they do. “I think that everybody has to be moved by these stories, and by the light that advances in technology have brought into the lives of families and kids and all of that. I don’t want to ever diminish that.” But he also said that it’s important to keep kids safe too. “Overall, it’s a public safety kind of issue.”

Take Jack’s hand, for example. It was broken out of the box, and Erenstone spent Christmas Eve rushing to fix it the best he could. “I super-glued the thing back together as best I could. But I knew how easily it broke and I knew it wasn’t going to last.” When Jack came in the next day to get the hand fitted, it broke again. Erenstone was able to get it working, but it broke when Jack got home as well.

Chaput, the volunteer who made the hand, said it was the first he ever printed for someone (to become an e-NABLE printer, volunteers have to print and assemble a test hand, but this was his first that a human would use). Chaput is a chemical engineer by day, and like the rest of the e-NABLE printers, he does all this work for free. He thinks two things probably went wrong in the printing, and both are endemic to the way that 3D printing works.

You can think of 3D printing like a very precise hot glue gun that lays down thin layers of hot plastic. This means that the pieces that get printed are very strong in some ways, but weak in others. So if you pull up on the piece, pulling perpendicular the direction the layers were laid down, it can break. This is how the biggest crack in Jack’s hand formed. The other, smaller cracks were likely due to another common 3D printing challenge: temperature.

“You’re always battling the temperature,” Chaput said. “When you extrude, you want it to come out soft obviously, it has to leave the nozzle and bond, but then you want it to harden quickly. It’s the soft but hard concept that you’re always battling.” Chaput said that he thinks that Jack’s hand was made a little bit too cold, which caused cracks to form.

For Chaput, this whole thing was a learning experience. “Every time you make one it comes out better. And that’s the thing, that whole hand was only eight bucks worth of plastic, so making more of them is no big deal.”

When I asked him if he was worried about sending something that might be broken to a kid to use, something that a kid could get hurt using, he said he was. “That’s why I like having Jeff [Erenstone] there. Sending it out to a totally random person that you don’t know what they’re going to do with it, particularly when they have a really young kid—that is an unsettling thought.”

But many of the e-NABLE volunteers do just that—they mail the printed hand to the person who asked for it. In the vast majority of cases, that’s fine. Since most kids aren’t using them for sports or intense activity they’re not likely to hurt themselves. And e-NABLE is careful to explain to recipients what the hands are capable of. But not everyone is like e-NABLE. There are other groups and companies advertising 3D printing as a full replacement for a hand. And that’s where Erenstone and Fries start to get worried. “3D printing does not break physics,” said Erenstone. Plastic can only take so much.

Jack’s story has an interesting coda, one that points to the future of 3D printed prosthetic devices. After his first Raptor hand came out of the box broken, Erenstone decided he would make something else for him. Something better. So he teamed up with Steve Wood, an engineer based in the UK who had become involved in the e-NABLE community and whose designs Erenstone described to me as “brilliant” more than once. In 2013, Wood came across a material called Filaflex—a more flexible material than the usual hard plastic. He started playing around with it. “I created a hinge between two rigid parts, and that grew into a finger because a finger is full of three hinges, and the finger then developed into a hand because I needed something to connect the finger too.” Eventually he had something he called a “Flexy-Hand.”

That was what Erenstone wanted to give Jack—so he sent scans of Jack’s hand to Wood and asked if he could make him one. Not only did Wood make a Flexy-Hand for him, he also printed out a copy of Jack’s hand to test the device out on. He sent both to Erenstone, and in January the Reidy family gathered in Erenstone’s office, with Wood on video chat, to test out the hand.

[EMBED” https://www.youtube.com/watch?v=9EocIKpdPyw]

Within a few minutes, Jack was picking up bottles, grasping cans, and even writing his name with his left hand—something he had never done before. “Think of the dexterity it takes to write your name. He’d never done that with his left hand before, because it wasn’t a possibility,” Erenstone said. Wood had never watched someone put on one of his devices for the first time. “He took to it like a fish to water,” Wood said.

But the Flexy-Hand isn’t quite the same as what the average e-NABLE volunteer is able to make. Wood is an engineer by training with years of experience in building and designing mechanical devices using special design software called CAD. “I’m sure I have a massive advantage in understanding CAD and having 28 years or so of engineering experience behind me. It must count for something.” And Filaflex isn’t easy to print, nor is it as cheap as the standard plastic. Not all 3D printers can handle the material, and it can be finicky.

Erenstone said that all told, including his time helping Jack fit the device, the Flexy-Hand probably costs $2,000. Compared to the standard 3D printed hand that’s a lot. But compared to a carbon fiber hand that might run something like $8,000, it’s not. And Erenstone said this was the first 3D printed prosthetic that he would be willing to put on a patient as a real prosthetic device.

But this is where the promising future of 3D printed hands probably lies. Not in the $30, volunteer-printed version, but in this middle ground where engineers and prosthetists work together to make something slightly cheaper than the average professionally made device.

Wood said he couldn’t make the hand without Erenstone’s help. “I can make custom designs made to measure all day long, but I’m not medically trained and I don’t have the qualifications for the fitting of prosthetics. This is I think where it becomes a good partnership between myself and Jeff.”

Jack has had his Flexy-Hand at home for about two months now. James said he was hesitant to use it at first since he didn’t want to break it like the earlier Raptor hand, but in the past few days Jack has started wearing it more. But even the fancy new hand doesn’t work for a lot of situations. On Thursday he tried the hand at hockey practice for the first time. It didn’t fit quite right in the glove, so he couldn’t use it. He also tried to shoot hoops with the hand, and he took it off pretty quickly. “With Jack it might have been different if he lost his hand after birth,” James said. “I think that he is so used to being without, especially with sports.”

Despite all the back and forth, James is hesitant to criticize the e-NABLE process. “I wouldn’t call them issues, since they’re just getting started,” he said. “It’s a great thing, but it’s not 100 percent functional for everything you do in life. I don’t want to knock it, it’s been great.”

Here’s how Schull thinks about 3D printed hands: “What I say these days is that these devices are compared favorably, especially by kids, to commercial prosthetics costing thousands or tens of thousands of dollars. They’re compared favorably. But, a 9-year-old will compare peanut butter very favorably to caviar. And indeed peanut butter is probably a better fit for that kid, but they’re just not the same.”

THEDAILYBEAST.COM
by Rose Eveleth | 03.11.15 5:15 AM ET

Smart robots thanks of 3D printing

3D Printing may be the key for robots to become capable of self-replicating, as well as self-learning! Evolving robots may be round the corner!

Self-replicating, evolving robots. We’re not too sure how we feel about that. 😐

http://www.ign.com/…/robots-use-3d-printing-and-simulations…

Researchers at Oslo University have combined computer simulations and 3D printers to improve robots through the process of evolution.

The goal, outlined in Apollon research magazine, is for future robots to be self-learning and self-replicating, armed with a 3D printer that would allow automatons to create new robots or tools to solve problems, like navigating an asteroid.

The concept is pretty simple – the team creates an obstacle course and gives the requirements to the computer, which produces several models for testing through simulation. The robots are then built using a 3D printer and tested in the real-world.

Right now the real-world results are only half as good as the computer simulations suggest but, by using this data to improve the algorithm, scientists hope future versions could be accurate enough for the robots to perform better in real-life than in a rigorous simulation.

The researchers claim that getting this right could be a major step forward in any number of applications, ranging from interplanetary mining to repairing damaged nuclear reactors.

IGN.COM
by SETH TIPPS | 13 NOV 2014