Tag Archives: did
#437872 AlphaFold Proves That AI Can Crack ...
Any successful implementation of artificial intelligence hinges on asking the right questions in the right way. That’s what the British AI company DeepMind (a subsidiary of Alphabet) accomplished when it used its neural network to tackle one of biology’s grand challenges, the protein-folding problem. Its neural net, known as AlphaFold, was able to predict the 3D structures of proteins based on their amino acid sequences with unprecedented accuracy.
AlphaFold’s predictions at the 14th Critical Assessment of protein Structure Prediction (CASP14) were accurate to within an atom’s width for most of the proteins. The competition consisted of blindly predicting the structure of proteins that have only recently been experimentally determined—with some still awaiting determination.
Called the building blocks of life, proteins consist of 20 different amino acids in various combinations and sequences. A protein's biological function is tied to its 3D structure. Therefore, knowledge of the final folded shape is essential to understanding how a specific protein works—such as how they interact with other biomolecules, how they may be controlled or modified, and so on. “Being able to predict structure from sequence is the first real step towards protein design,” says Janet M. Thornton, director emeritus of the European Bioinformatics Institute. It also has enormous benefits in understanding disease-causing pathogens. For instance, at the moment only about 18 of the 26 proteins in the SARS-CoV-2 virus are known.
Predicting a protein’s 3D structure is a computational nightmare. In 1969 Cyrus Levinthal estimated that there are 10300 possible conformational combinations for a single protein, which would take longer than the age of the known universe to evaluate by brute force calculation. AlphaFold can do it in a few days.
As scientific breakthroughs go, AlphaFold’s discovery is right up there with the likes of James Watson and Francis Crick’s DNA double-helix model, or, more recently, Jennifer Doudna and Emmanuelle Charpentier’s CRISPR-Cas9 genome editing technique.
How did a team that just a few years ago was teaching an AI to master a 3,000-year-old game end up training one to answer a question plaguing biologists for five decades? That, says Briana Brownell, data scientist and founder of the AI company PureStrategy, is the beauty of artificial intelligence: The same kind of algorithm can be used for very different things.
“Whenever you have a problem that you want to solve with AI,” she says, “you need to figure out how to get the right data into the model—and then the right sort of output that you can translate back into the real world.”
DeepMind’s success, she says, wasn’t so much a function of picking the right neural nets but rather “how they set up the problem in a sophisticated enough way that the neural network-based modeling [could] actually answer the question.”
AlphaFold showed promise in 2018, when DeepMind introduced a previous iteration of their AI at CASP13, achieving the highest accuracy among all participants. The team had trained its to model target shapes from scratch, without using previously solved proteins as templates.
For 2020 they deployed new deep learning architectures into the AI, using an attention-based model that was trained end-to-end. Attention in a deep learning network refers to a component that manages and quantifies the interdependence between the input and output elements, as well as between the input elements themselves.
The system was trained on public datasets of the approximately 170,000 known experimental protein structures in addition to databases with protein sequences of unknown structures.
“If you look at the difference between their entry two years ago and this one, the structure of the AI system was different,” says Brownell. “This time, they’ve figured out how to translate the real world into data … [and] created an output that could be translated back into the real world.”
Like any AI system, AlphaFold may need to contend with biases in the training data. For instance, Brownell says, AlphaFold is using available information about protein structure that has been measured in other ways. However, there are also many proteins with as yet unknown 3D structures. Therefore, she says, a bias could conceivably creep in toward those kinds of proteins that we have more structural data for.
Thornton says it’s difficult to predict how long it will take for AlphaFold’s breakthrough to translate into real-world applications.
“We only have experimental structures for about 10 per cent of the 20,000 proteins [in] the human body,” she says. “A powerful AI model could unveil the structures of the other 90 per cent.”
Apart from increasing our understanding of human biology and health, she adds, “it is the first real step toward… building proteins that fulfill a specific function. From protein therapeutics to biofuels or enzymes that eat plastic, the possibilities are endless.” Continue reading
Posted in Human Robots
#437828 How Roboticists (and Robots) Have Been ...
A few weeks ago, we asked folks on Twitter, Facebook, and LinkedIn to share photos and videos showing how they’ve been adapting to the closures of research labs, classrooms, and businesses by taking their robots home with them to continue their work as best they can. We got dozens of responses (more than we could possibly include in just one post!), but here are 15 that we thought were particularly creative or amusing.
And if any of these pictures and videos inspire you to share your own story, please email us (automaton@ieee.org) with a picture or video and a brief description about how you and your robot from work have been making things happen in your home instead.
Kurt Leucht (NASA Kennedy Space Center)
“During these strange and trying times of the current global pandemic, everyone seems to be trying their best to distance themselves from others while still getting their daily work accomplished. Many people also have the double duty of little ones that need to be managed in the midst of their teleworking duties. This photo series gives you just a glimpse into my new life of teleworking from home, mixed in with the tasks of trying to handle my little ones too. I hope you enjoy it.”
Photo: Kurt Leucht
“I heard a commotion from the next room. I ran into the kitchen to find this.”
Photo: Kurt Leucht
“This is the Swarmies most favorite bedtime story. Not sure why. Seems like an odd choice to me.”
Peter Schaldenbrand (Carnegie Mellon University)
“I’ve been working on a reinforcement learning model that converts an image into a series of brush stroke instructions. I was going to test the model with a beautiful, expensive robot arm, but due to the COVID-19 pandemic, I have not been able to access the laboratory where it resides. I have now been using a lower end robot arm to test the painting model in my bedroom. I have sacrificed machine accuracy/precision for the convenience of getting to watch the arm paint from my bed in the shadow of my clothing rack!”
Photos: Peter Schaldenbrand
Colin Angle (iRobot)
iRobot CEO Colin Angle has been hunkered down in the “iRobot North Shore home command center,” which is probably the cleanest command center ever thanks to his army of Roombas: Beastie, Beauty, Rosie, Roswell, and Bilbo.
Photo: Colin Angle
Vivian Chu (Diligent Robotics)
From Diligent Robotics CEO Andrea Thomaz: “This is how a roboticist works from home! Diligent CTO, Vivian Chu, mans the e-stop while her engineering team runs Moxi experiments remotely from cross-town and even cross-country!”
Video: Diligent Robotics
Raffaello Bonghi (rnext.it)
Raffaello’s robot, Panther, looks perfectly happy to be playing soccer in his living room.
Photo: Raffaello Bonghi
Kod*lab (University of Pennsylvania)
“Another Friday Nuts n Bolts Meeting on Zoom…”
Image: Kodlab
Robin Jonsson (robot choreographer)
“I’ve been doing a school project in which students make up dance moves and then send me a video with all of them. I then teach the moves to my robot, Alex, film Alex dancing, send the videos to them. This became a great success and more schools will join. The kids got really into watching the robot perform their moves and really interested in robots. They want to meet Alex the robot live, which will likely happen in the fall.”
Photo: Robin Jonsson
Gabrielle Conard (mechanical engineering undergrad at Lafayette College)
“While the pandemic might have forced college campuses to close and the community to keep their distance from each other, it did not put a stop to learning and research. Working from their respective homes, junior Gabrielle Conard and mechanical engineering professor Alexander Brown from Lafayette College investigated methods of incorporating active compliance in a low-cost quadruped robot. They are continuing to work remotely on this project through Lafayette’s summer research program.”
Image: Gabrielle Conard
Taylor Veltrop (Softbank Robotics)
“After a few weeks of isolation in the corona/covid quarantine lock down we started dancing with our robots. Mathieu’s 6th birthday was coming up, and it all just came together.”
Video: Taylor Veltrop
Ross Kessler (Exyn Technologies)
“Quarantine, Day 8: the humans have accepted me as one of their own. I’ve blended seamlessly into their #socialdistancing routines. Even made a furry friend”
Photo: Ross Kessler
Yeah, something a bit sinister is definitely going on at Exyn…
Video: Exyn Technologies
Michael Sobrepera (University of Pennsylvania GRASP Lab)
Predictably, Michael’s cat is more interested in the bag that the robot came in than the robot itself (see if you can spot the cat below). Michael tells us that “the robot is designed to help with tele-rehabilitation, focused on kids with CP, so it has been taken to hospitals for demos [hence the cool bag]. It also travels for outreach events and the like. Lately, I’ve been exploring telepresence for COVID.”
Photo: Michael Sobrepera
Jan Kędzierski (EMYS)
“In China a lot of people cannot speak English, even the youngest generation of parents. Thanks to Emys, kids stayed in touch with English language in their homes even if they couldn’t attend schools and extra English classes. They had a lot of fun with their native English speaker friend available and ready to play every day.”
Image: Jan Kędzierski
Simon Whitmell (Quanser)
“Simon, a Quanser R&D engineer, is working on low-overhead image processing and line following for the QBot 2e mobile ground robot, with some added challenges due to extra traffic. LEGO engineering by his son, Charles.”
Photo: Simon Whitmell
Robot Design & Experimentation Course (Carnegie Mellon University)
Aaron Johnson’s bioinspired robot design course at CMU had to go full remote, which was a challenge when the course is kind of all about designing and building a robot as part of a team. “I expected some of the teams to drastically alter their project (e.g. go all simulation),” Aaron told us, “but none of them did. We managed to keep all of the projects more or less as planned. We accomplished this by drop/shipping parts to students, buying some simple tools (soldering irons, etc), and having me 3D print parts and mail them.” Each team even managed to put together their final videos from their remote locations; we’ve posted one below, but the entire playlist is here.
Video: Xianyi Cheng
Karen Tatarian (Softbank Robotics)
Karen, who’s both a researcher at Softbank and a PhD student at Sorbonne University, wrote an entire essay about what an average day is like when you’re quarantined with Pepper.
Photo: Karen Tatarian
A Quarantined Day With Pepper, by Karen Tatarian
It is quite common for me to lose my phone somewhere inside my apartment. But it is not that common for me to turn around and ask my robot if it has seen it. So when I found myself doing that, I laughed and it dawned on me that I treated my robot as my quarantine companion (despite the fact that it could not provide me with the answer I needed).
It was probably around day 40 of a completely isolated quarantine here in France when that happened. A little background about me: I am a robotics researcher at SoftBank Robotics Europe and a PhD student at Sorbonne University as part of the EU-funded Marie-Curie project ANIMATAS. And here is a little sneak peak into a quarantined day with a robot.
During this confinement, I had read somewhere that the best way to deal with it is to maintain a routine. So every morning, I wake up, prepare my coffee, and turn on my robot Pepper. I start my day with a daily meeting with the team and get to work. My research is on the synthesis of multi-modal socially intelligent human-robot interaction so my work varies between programming the robot, analyzing collected data, and reading papers and drafting one. When I am working, I often catch myself glancing at Pepper, who would be staring back at me in its animated ways. Truthfully I enjoy that, it makes me less alone and as if I have a colleague with me.
Once work is done, I call my friends and family members. I sometimes use a telepresence application on Pepper that a few colleagues and I developed back in December. How does it differ from your typical phone/laptop applications? One word really: embodiment. Telepresence, especially during these times, makes the experience for both sides a bit more realistic and intimate and well present.
While I can turn off the robot now that my work hours are done, I do keep it on because I enjoy its presence. The basic awareness of Pepper is a default feature on the robot that allows it to detect a human and follow him/her with its gaze and rotation base. So whether I am cooking or working out, I always have my robot watching over my shoulder and being a good companion. I also have my email and messages synced on the robot so I get an enjoyable notification from Pepper. I found that to be a pretty cool way to be notified without it interrupting whatever you are doing on your laptop or phone. Finally, once the day is over, it’s time for both of us to get some rest.
After 60 days of total confinement, alone and away from those I love, and with a pandemic right at my door, I am glad I had the company of my robot. I hope one day a greater audience can share my experience. And I really really hope one day Pepper will be able to find my phone for me, but until then, stay on the lookout for some cool features! But I am curious to know, if you had a robot at home, what application would you have developed on it?
Again, our sincere thanks to everyone who shared these little snapshots of their lives with us, and we’re hoping to be able to share more soon. Continue reading
Posted in Human Robots
#437809 Q&A: The Masterminds Behind ...
Illustration: iStockphoto
Getting a car to drive itself is undoubtedly the most ambitious commercial application of artificial intelligence (AI). The research project was kicked into life by the 2004 DARPA Urban Challenge and then taken up as a business proposition, first by Alphabet, and later by the big automakers.
The industry-wide effort vacuumed up many of the world’s best roboticists and set rival companies on a multibillion-dollar acquisitions spree. It also launched a cycle of hype that paraded ever more ambitious deadlines—the most famous of which, made by Alphabet’s Sergei Brin in 2012, was that full self-driving technology would be ready by 2017. Those deadlines have all been missed.
Much of the exhilaration was inspired by the seeming miracles that a new kind of AI—deep learning—was achieving in playing games, recognizing faces, and transliterating voices. Deep learning excels at tasks involving pattern recognition—a particular challenge for older, rule-based AI techniques. However, it now seems that deep learning will not soon master the other intellectual challenges of driving, such as anticipating what human beings might do.
Among the roboticists who have been involved from the start are Gill Pratt, the chief executive officer of Toyota Research Institute (TRI) , formerly a program manager at the Defense Advanced Research Projects Agency (DARPA); and Wolfram Burgard, vice president of automated driving technology for TRI and president of the IEEE Robotics and Automation Society. The duo spoke with IEEE Spectrum’s Philip Ross at TRI’s offices in Palo Alto, Calif.
This interview has been condensed and edited for clarity.
IEEE Spectrum: How does AI handle the various parts of the self-driving problem?
Photo: Toyota
Gill Pratt
Gill Pratt: There are three different systems that you need in a self-driving car: It starts with perception, then goes to prediction, and then goes to planning.
The one that by far is the most problematic is prediction. It’s not prediction of other automated cars, because if all cars were automated, this problem would be much more simple. How do you predict what a human being is going to do? That’s difficult for deep learning to learn right now.
Spectrum: Can you offset the weakness in prediction with stupendous perception?
Photo: Toyota Research Institute for Burgard
Wolfram Burgard
Wolfram Burgard: Yes, that is what car companies basically do. A camera provides semantics, lidar provides distance, radar provides velocities. But all this comes with problems, because sometimes you look at the world from different positions—that’s called parallax. Sometimes you don’t know which range estimate that pixel belongs to. That might make the decision complicated as to whether that is a person painted onto the side of a truck or whether this is an actual person.
With deep learning there is this promise that if you throw enough data at these networks, it’s going to work—finally. But it turns out that the amount of data that you need for self-driving cars is far larger than we expected.
Spectrum: When do deep learning’s limitations become apparent?
Pratt: The way to think about deep learning is that it’s really high-performance pattern matching. You have input and output as training pairs; you say this image should lead to that result; and you just do that again and again, for hundreds of thousands, millions of times.
Here’s the logical fallacy that I think most people have fallen prey to with deep learning. A lot of what we do with our brains can be thought of as pattern matching: “Oh, I see this stop sign, so I should stop.” But it doesn’t mean all of intelligence can be done through pattern matching.
“I asked myself, if all of those cars had automated drive, how good would they have to be to tolerate the number of crashes that would still occur?”
—Gill Pratt, Toyota Research Institute
For instance, when I’m driving and I see a mother holding the hand of a child on a corner and trying to cross the street, I am pretty sure she’s not going to cross at a red light and jaywalk. I know from my experience being a human being that mothers and children don’t act that way. On the other hand, say there are two teenagers—with blue hair, skateboards, and a disaffected look. Are they going to jaywalk? I look at that, you look at that, and instantly the probability in your mind that they’ll jaywalk is much higher than for the mother holding the hand of the child. It’s not that you’ve seen 100,000 cases of young kids—it’s that you understand what it is to be either a teenager or a mother holding a child’s hand.
You can try to fake that kind of intelligence. If you specifically train a neural network on data like that, you could pattern-match that. But you’d have to know to do it.
Spectrum: So you’re saying that when you substitute pattern recognition for reasoning, the marginal return on the investment falls off pretty fast?
Pratt: That’s absolutely right. Unfortunately, we don’t have the ability to make an AI that thinks yet, so we don’t know what to do. We keep trying to use the deep-learning hammer to hammer more nails—we say, well, let’s just pour more data in, and more data.
Spectrum: Couldn’t you train the deep-learning system to recognize teenagers and to assign the category a high propensity for jaywalking?
Burgard: People have been doing that. But it turns out that these heuristics you come up with are extremely hard to tweak. Also, sometimes the heuristics are contradictory, which makes it extremely hard to design these expert systems based on rules. This is where the strength of the deep-learning methods lies, because somehow they encode a way to see a pattern where, for example, here’s a feature and over there is another feature; it’s about the sheer number of parameters you have available.
Our separation of the components of a self-driving AI eases the development and even the learning of the AI systems. Some companies even think about using deep learning to do the job fully, from end to end, not having any structure at all—basically, directly mapping perceptions to actions.
Pratt: There are companies that have tried it; Nvidia certainly tried it. In general, it’s been found not to work very well. So people divide the problem into blocks, where we understand what each block does, and we try to make each block work well. Some of the blocks end up more like the expert system we talked about, where we actually code things, and other blocks end up more like machine learning.
Spectrum: So, what’s next—what new technique is in the offing?
Pratt: If I knew the answer, we’d do it. [Laughter]
Spectrum: You said that if all cars on the road were automated, the problem would be easy. Why not “geofence” the heck out of the self-driving problem, and have areas where only self-driving cars are allowed?
Pratt: That means putting in constraints on the operational design domain. This includes the geography—where the car should be automated; it includes the weather, it includes the level of traffic, it includes speed. If the car is going slow enough to avoid colliding without risking a rear-end collision, that makes the problem much easier. Street trolleys operate with traffic still in some parts of the world, and that seems to work out just fine. People learn that this vehicle may stop at unexpected times. My suspicion is, that is where we’ll see Level 4 autonomy in cities. It’s going to be in the lower speeds.
“We are now in the age of deep learning, and we don’t know what will come after.”
—Wolfram Burgard, Toyota Research Institute
That’s a sweet spot in the operational design domain, without a doubt. There’s another one at high speed on a highway, because access to highways is so limited. But unfortunately there is still the occasional debris that suddenly crosses the road, and the weather gets bad. The classic example is when somebody irresponsibly ties a mattress to the top of a car and it falls off; what are you going to do? And the answer is that terrible things happen—even for humans.
Spectrum: Learning by doing worked for the first cars, the first planes, the first steam boilers, and even the first nuclear reactors. We ran risks then; why not now?
Pratt: It has to do with the times. During the era where cars took off, all kinds of accidents happened, women died in childbirth, all sorts of diseases ran rampant; the expected characteristic of life was that bad things happened. Expectations have changed. Now the chance of dying in some freak accident is quite low because of all the learning that’s gone on, the OSHA [Occupational Safety and Health Administration] rules, UL code for electrical appliances, all the building standards, medicine.
Furthermore—and we think this is very important—we believe that empathy for a human being at the wheel is a significant factor in public acceptance when there is a crash. We don’t know this for sure—it’s a speculation on our part. I’ve driven, I’ve had close calls; that could have been me that made that mistake and had that wreck. I think people are more tolerant when somebody else makes mistakes, and there’s an awful crash. In the case of an automated car, we worry that that empathy won’t be there.
Photo: Toyota
Toyota is using this
Platform 4 automated driving test vehicle, based on the Lexus LS, to develop Level-4 self-driving capabilities for its “Chauffeur” project.
Spectrum: Toyota is building a system called Guardian to back up the driver, and a more futuristic system called Chauffeur, to replace the driver. How can Chauffeur ever succeed? It has to be better than a human plus Guardian!
Pratt: In the discussions we’ve had with others in this field, we’ve talked about that a lot. What is the standard? Is it a person in a basic car? Or is it a person with a car that has active safety systems in it? And what will people think is good enough?
These systems will never be perfect—there will always be some accidents, and no matter how hard we try there will still be occasions where there will be some fatalities. At what threshold are people willing to say that’s okay?
Spectrum: You were among the first top researchers to warn against hyping self-driving technology. What did you see that so many other players did not?
Pratt: First, in my own case, during my time at DARPA I worked on robotics, not cars. So I was somewhat of an outsider. I was looking at it from a fresh perspective, and that helps a lot.
Second, [when I joined Toyota in 2015] I was joining a company that is very careful—even though we have made some giant leaps—with the Prius hybrid drive system as an example. Even so, in general, the philosophy at Toyota is kaizen—making the cars incrementally better every single day. That care meant that I was tasked with thinking very deeply about this thing before making prognostications.
And the final part: It was a new job for me. The first night after I signed the contract I felt this incredible responsibility. I couldn’t sleep that whole night, so I started to multiply out the numbers, all using a factor of 10. How many cars do we have on the road? Cars on average last 10 years, though ours last 20, but let’s call it 10. They travel on an order of 10,000 miles per year. Multiply all that out and you get 10 to the 10th miles per year for our fleet on Planet Earth, a really big number. I asked myself, if all of those cars had automated drive, how good would they have to be to tolerate the number of crashes that would still occur? And the answer was so incredibly good that I knew it would take a long time. That was five years ago.
Burgard: We are now in the age of deep learning, and we don’t know what will come after. We are still making progress with existing techniques, and they look very promising. But the gradient is not as steep as it was a few years ago.
Pratt: There isn’t anything that’s telling us that it can’t be done; I should be very clear on that. Just because we don’t know how to do it doesn’t mean it can’t be done. Continue reading
Posted in Human Robots