Robots Beat Human Records At Beijing Half-Marathon (techcrunch.com) 74
An anonymous reader quotes a report from TechCrunch: The winning runner at a Beijing half-marathon for humanoid robots finished the race today in 50 minutes and 26 seconds -- significantly faster than the human world record of 57 minutes recently set by Jacob Kiplimo. [...] [T]he winning time is a massive improvement over last year's race, when the fastest robot finished in two hours and 40 minutes.
The Associated Press reports that this year's winner was built by Chinese smartphone maker Honor. It seems the winning robot wasn't actually the fastest, as a different Honor robot finished in 48 minutes and 19 seconds. But that one was remote controlled -- the 50:26 robot was autonomous and won due to weighted scoring. About 40% of participating robots competed autonomously, while the remaining 60% were remote controlled, according to Beijing's E-Town tech hub. Not all of them did as well as Honor's robots, with one robot falling at the starting line and another hitting a barrier.
The Associated Press reports that this year's winner was built by Chinese smartphone maker Honor. It seems the winning robot wasn't actually the fastest, as a different Honor robot finished in 48 minutes and 19 seconds. But that one was remote controlled -- the 50:26 robot was autonomous and won due to weighted scoring. About 40% of participating robots competed autonomously, while the remaining 60% were remote controlled, according to Beijing's E-Town tech hub. Not all of them did as well as Honor's robots, with one robot falling at the starting line and another hitting a barrier.
Um...so what? (Score:1, Redundant)
Re:Um...so what? (Score:4, Insightful)
To me, this is more about the state of the art of robots, than a comparison to human capabilities.
Until now, robots have *not* been able to accomplish such a feat. That in itself makes it noteworthy.
Can machines go faster? Sure. But the robot needed to keep its balance the entire time, and not run out of battery power. These have not been easy challenges for robot designers to overcome.
Re: (Score:2, Insightful)
Machines have been faster for a LONG time when they're not arbitrarily constrained by being forced to emulate animal motion.
Machines which fly or run on wheels are far more efficient at traversing fixed distances at speed, and have long existed.
Re:Um...so what? (Score:5, Insightful)
Of course you're right, when they're not arbitrarily constrained by being forced to emulate animal motion
Whether you think this is a reasonable constraint or not, it's still an achievement.
There are good reasons to make machines emulate animal motions. For example, a wheeled indoor delivery robot can't climb stairs. A humanoid robot would be able to do this much better. A car can go much faster than a human, but it can't go through the woods where there are no paths, it needs roads.
So every kind of machine has constraints of some sort, just different constraints. If there is no use case for a humanoid robot that can run marathons, it won't sell, and the concept will die. But we will still have learned a lot about how to make robots balance, and operate efficiently, in the process.
Re: (Score:2)
I hate to tell you that "animal motion" is pretty damn good. We've had several million years to perfect it. Now the knee... that needs some work. Just not designed for longevity.
Re: (Score:2)
lets see if the robot lasts the 60+years of use that knees tend to do.
Re: (Score:2)
Of course, longevity will be an area where a lot of improvement will be needed. In the early 20th century, cars and planes didn't last long either.
Re: (Score:2)
OK... But why would it need to? I mean, human joints do wear out. When critical ones do, the standard medical practice today is to replace them. You can do exactly the same thing with robots, but it's generally far easier. An efficient system would be modular. That is to say that the part that wears out might be some sort of spring, but the best thing to do would be to take out the entire joint, pop in a replacement and send the joint module back to a central factory to be reconditioned. That would allow fo
Re: Um...so what? (Score:2)
On the contrary , the knee is perfectly evolved to last the *natural* human lifespan, ie about 40-50 years.
Re: (Score:2)
Indeed, there are whole categories of human inventions that focus on imitation of human or animal capabilities, from hearing to seeing to speaking to flying to swimming to thinking.
You've got a point about knees!
Re: (Score:2)
There are good reasons to make machines emulate animal motions. For example, a wheeled indoor delivery robot can't climb stairs
Not sure why not. It's pretty common now for delivery to use powered hand trucks that do the stair climbing themselves. Some of them have tracks, but some of them just use wheels. Take out the human operator and you have a wheeled indoor delivery robot that can climb stairs. Various types of stair climbing load carriers, including wheelchairs have been around for something like 70 years now.
Maybe feet would be better, but why not wheeled feet? They could go through woods just fine as well, not to mention pr
Re:Um...so what? (Score:5, Insightful)
If you absolutely insist on practical applications, how about better prosthetics? Powered "exoskeletons" for paralyzed people or those with severe neuromuscular disorders? Or (and sadly much more likely) combat robots for terrain like dense jungle where neither wheels nor flying works?
The inventor of the transistor was convinced that it had no practical purpose whatsoever.
Re: (Score:2)
Pretty sure you can fly over every jungle that has ever existed.
Re: (Score:2)
Re: Um...so what? (Score:2)
Could use IR cameras now, dont need defolliant.
Re: (Score:2)
Put it to you this way: You know what else can beat a human in a marathon? A car.
Re: (Score:3)
Re: (Score:2)
But what is the use? This isn't a bending unit from Futurama. This is finally accomplishing the first of many rudimentary, basic tasks. This kind of generic robot suffers from an enormous "80-20" "long tail" problem, and they aren't even close to the 80% mark yet.
When one of these things can scale Mt. Everest unassisted, I'll say they have something to celebrate.
Re: (Score:2)
Who knows what the uses could be! There are a billion places cars or airplanes can't go, that a humanoid robot can. How about, up a flight of stairs, for example. A humanoid robot would need far fewer special accommodations than other types of locomotion, especially indoors. If the only thing the robot can do is carry a thing from one place to another, in some contexts that's already amazing.
We've only just begun the development of humanoid robots. Of course there are many problems to solve. There is plenty
Re:Um...so what? (Score:4, Informative)
Machine faster than human. They may also be physically stronger. The only thing this shows, is that humanoid robots have become reasonably efficient (assuming no recharge breaks).
I am so puzzled by this kind of reaction. This is the first time something like that has ever happened! Check out "WABOT-1" -- I'm not sure if it was the first bipedal robot to walk at all, but it's a product of the 1970s. Look at the progress in 50 years!
Then Check out Honda Asimo from ~2010. Asimo was only about 15 years ago.
I think going from WABOT to Asimo to an autonomous running robot that can run 26 miles in under 2 hours—in half a century—is absolutely amazing. The engineers who designed and built it should be applauded.
Re: (Score:2)
Excuse me, my error. Half marathon, 13 miles.
Re:Um...so what? (Score:4, Insightful)
This is cool stuff. It's a big advancement in the technology. It's 100% a nerd thing, on a nerd site. But no. The reactions we get is akin to saying "my horse can go as fast as your steam engine contraption" back in the 1880s. These robots are much better than humanoid robots from last year, nevermind from 5 or 10 years ago. Even if you think it's kind of pointless now, in another 5 years? Our wars will be different, but also maybe food delivery will go to robots along with a lot of other manual labor that is currently not automatable. You can argue that's concerning for the future, and on that I'd agree, but speaking as a nerd, it's still fucking cool.
Re: (Score:2)
You can argue that's concerning for the future, and on that I'd agree, but speaking as a nerd, it's still fucking cool.
It kind of feels to me like it's hard to get excited about things these days? I mean that if you think a particular breakthrough is cool, well you didn't consider problems X or Y. And it was done better, by Z. And it's going to destroy the environment. And people on the opposite side of the political spectrum like it, so that's a problem too. Etc.
I'm not sure if we're really in an unusually negative-thinking period if history, or if it just feels that way to me, in the current political climate, in my own s
Re: (Score:2)
Human extinction. Is that enough so-what for you? (Score:2)
Actually the frightening part of the story is that the fastest was remotely operated. Humans today, but operated by a malignant ASI tomorrow. Well, hopefully not tomorrow. I'd prefer not to see the end of this story and I'm hoping to be around tomorrow and even for a few more years. But RSN?
Too many books could be cited, but it's not like today's Slashdotters seem to have much interest in books. Can't resist a recent one with high relevance to this story: Army of None by Paul Scharre about autonomous weap
Re: (Score:2)
Too many books could be cited, but it's not like today's Slashdotters seem to have much interest in books.
If books aren't your speed then how about the somewhat okayish Bruce Willis action flick Surrogates [wikipedia.org]? A 2009 film that asks the question, "What happens when every human in the world lives their entire life through the eyes of a human operated android?"
Re: (Score:2)
I don't watch many movies. Never watched many, and far fewer lately. However I am unable to recall an example of a movie that I thought was better than the book. Most often I felt like the movie eliminated many of the imaginative possibilities of the book. Largely a matter of bandwidth? Movies flood the zone, filling both the visual and audio channels and requiring almost all of your mental capacity to keep up. More so as the effects have become more special and dazzling. For books you have to do most of th
Re: Um...so what? (Score:2)
Well, up until now, robots havenâ(TM)t been faster than humans at this task. In fact, robots have been catastrophically bad at it. Itâ(TM)s a demonstration that bipedal robot technology is advancing rapidly.
Re: (Score:2)
I bet a one-wheel, battery-powered skateboard can beat it.
Re: (Score:2)
Re: (Score:2)
Makes me think of Short Circuit:
Ben Jabituya: "Unable. Malfunction."
Howard Marner: How can it refuse to turn itself off?
Shroeder: Maybe it's pissed off.
Newton Crosby: It's a machine, Schroeder. It doesn't get pissed off. It doesn't get happy, it doesn't get sad, it doesn't laugh at your jokes.
Newton Crosby, Ben Jabituya: [in unison] It just runs programs.
Howard Marner: It usually runs programs.
Video of robots in the race (Score:4, Funny)
Including the one that tripped ...
Robots in Beijing half-marathon - https://www.youtube.com/watch?... [youtube.com]
Re: (Score:2)
Obvious fake (Score:1)
Wut ?
Re: (Score:2)
Nah, I'm pretty sure this one's a robot. But it's a robot that's purpose-built to run a half-marathon. Huh? What's the utility of that? And somebody's going to have to convince me that it's not remote controlled.
Robot locomotion (Score:3)
The best concept I've seen involved a robot with knee and toe wheels. When presented with a smooth surface, it knelt and could travel at high speed. It only used legs when obstacles required the extra flexibility.
Re: (Score:2)
Indeed, arbitrarily constraining a robot to animal-like motion defeats many of the advantages of a robot. In most cases it's simply more efficient to use wheels or fly, which machines have long been capable of. It's only in a very few niche cases that walking on legs makes sense.
Re: (Score:2)
I agree that we shouldn't feel that we need to contain robot designs to humanoid or humanoid-esque shapes and mechanics, but I would disagree that there are only a very few niche cases where walking on legs make sense. Legs (limbs more generally) have millions of years of evolutions behind them, and they work very well in many situations (walking on the ground, running, climb trees, ascending a cliff, swimming, etc). They may be considered jacks-of-all trades, and they may be less efficient than wheels or t
Re: (Score:2)
That's because evolution is solving for a different problem. If organic "wheels" solved the problem, we would have them.
Re: (Score:2)
Wheels, at least macroscopic scales, are probably impossible for evolution to produce. Anything that has to freely rotate, in fact.
But it would be fun if that weren't true - I'd expect to see fish with propellers for tails.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
David Brin is a science fiction author who wrote a collection of books in the "Uplift" universe. The 30-second summary is that there are countless sentient species around the universe and a galactic civilization spanning billions of years. before humanity was discovered, no known example of evolution creating advanced intelligence was ever discovered. Instead, superior advanced species "uplifted" primitive species to be intelligent through gene modification, selective breeding, etc.
One of the species in the
Storming a defended beach... (Score:2)
Storming a defended beach may seem like a niche case for you but 90 years ago it had plenty of usage...
And if things go downhill, it still may...
Re: (Score:2)
Even 90 years ago attacking from the air formed a key part of the strategy. Today even more so.
You can see the battlefields of Ukraine where soldiers on legs are gradually being replaced with flying and wheeled/tracked robots.
The sort of ground defences that impede tracked/wheeled vehicles are easily able to impede bipedal attackers too. But machines which are able to fly above the battlefield are immune to these kind of defences.
Re: (Score:1)
In the near term, robot designs will likely benefit from being compatible with human infrastructure (escalators, stairs, ladders, doors, vehicles, etc). One strategy for doing that in a versatile way is to adopt human-like form and dynamics. It surprises me that designs can already exceed human capabilities in something as complex as running on real terrain, *while carrying their own power-supply*, for ~hours. As others have noted, an important take-away is the power-efficiency.
Re: (Score:2)
Being limited to a bipedal model is also pretty limiting. Animals with additional pairs of legs can navigate with much greater precision, ease, and efficiency.
Finally (Score:3)
I've never understood why athletes waste their time training and competing to get into peak physical fitness just so they can run around the block a few times. Fastest in the world, sure, but what's the point? Finally they can be replaced by robots that can do all that tedious work for them.
Maybe next they can work on a machine to pray for you, like a sort of electric monk? That would be a real time-saver for a lot of people.
Re: (Score:2)
- yea I don't want to hear about rain or snow or living in a shit hole, that's a you problem.
Re: (Score:2)
I've never understood why athletes waste their time training and competing to get into peak physical fitness just so they can run around the block a few times.
I don't get it either, but then I'm sure some of the tings I do in my life don't make sense to those pushing for peak physical fitness. Though this does remind me of a documentary I'd seen years ago, I really wish I could remember the name right now, where they'd discussed how humans had evolved over the millennia to run as part of our need to hunt for food.
Since humanity started in Africa the typical animals being hunted were large, nomadic, and often faster than humans so we needed to evolve to keep up.
Re: (Score:2)
This was W4 model. (Score:2)
It has a military T4 edition that can shoot as well..
How "big" is the robot? (Score:1)
The winning 'bot autonomously ran 25kph over a 21km course, that's a lot of computation to keep track of location while avoiding obstacles, especially at speed, but the article doesn't mention how tall / big this thing is. How much battery capacity did it have? Did it need to be recharged part way through the trip? What kind of processing did it have, was it on board or did it communicate via cell to an offsite server brute forcing the 'bot's movement?
I have to say it's an impressive feat of engineering
Re: (Score:2)
Re: (Score:2)
"Experts said the skills on display during the half-marathon, while entertaining, do not translate to the widespread commercialization of humanoid robots in industrial settings, where manual dexterity, real-world perception and capabilities beyond small-scale, repetitive tasks are crucial." which I think is crucial.
It's crucial at the moment but, as pointed out in other comments, we'd gone from a concept in the 70's to Asimo in the 2010s to a robot that can autonomously complete a half marathon in the span of 50 years. How long will it take researchers to improve the perception engine and energy storage before they're able to strap on an integrated sealed backpack and use them as short haul delivery 'bots in compounds / cities with lots of stairs and / or traffic congestion?
I'm not worried right now but I've got niec
Re: (Score:2)
Re: (Score:2)
Still a ways from Millie from the Jetsons handling household chores from cooking to washing dishes/clothes to vacuuming.
I believe you are think of Rosey, not Millie.
Re: (Score:2)
Oh, thank goodness. (Score:2)
Now robots can exercise for us! FINALLY! Something I'm willing to gladly hand over to our robot overlords!
Cool (Score:1)
Can you power the robot on rice and vegetables? Otherwise it seems less versatile if we're not comparing energy sources in a race.
I don't get to enter a car race running top fuel while everyone else is using a wood-fired steam boiler. And even if raced unofficially, just to prove something, I don't think anyone be impressed.
Re: (Score:1)
Re: (Score:2)
no, which is why I don't think I should be allowed to run on foot in an electric car race.
Change headline (Score:1)
The real story is all the robots lost the marathon because they can only run half of it. Humans would win, tortoise vs the hare.
Fighter jet? Millennium Falcon? (Score:1)
How fast could a fighter jet or a drone have done it? What about a fictional spaceship? Maybe the Millennium Falcon, eh?
All these options seem valid in the light of how utterly stupid this headline already is.
"Down in the Park", "Are Friends Electric?" (Score:2)
No fun (Score:2)
We'll miss all those articles, claims, excuses about doping, evading the doping tests and so on.
i wonder (Score:2)
I wonder how fast a Model T could have completed it. I also wonder if the robot can even complete a full marathon, considering needed weight of battery.
AI finally is a reality (Score:1)
Of course a machine can go faster than a human being. It can even fly, dive, go to space.