SoftBank Agrees to Buy Boston Dynamics From Alphabet
bloomberg.com
bloomberg.com
As I wrote last year when BD came up, the guy behind BigDog was Dr. Martin Buehler, who previously had an ambulatory robotics lab at McGill, and whose group built the first good running robot quadruped.[1] He's at Disney now. Raibert was primarily responsible for the hydraulics; his name is on the patent for that. But unless SoftBank wants to build giant mecha, hydraulic systems are too bulky. Also, Raibert is 68 now.
When Google bought both companies, I thought they'd put Boston Dynamics control algorithms on Schaft hardware. That did not happen, as far as anyone on the outside can tell.
The answer is that Alphabet doesn't see itself as a search company.
https://www.extremetech.com/extreme/185570-google-finally-pr...
In a wide sense Google is an artificial intelligence company. Search is an AI application. No wonder they try to enter the autonomous robots sector, first with cars, then with that bomb dog.
To continue to show growth, alphabet planned to bet a moderate investment on entering new sectors with huge growth prospects, making medium term investments by buying leading companies and giving them the funding they needed to get ahead.
This is why Google no longer does stuff like Google X, 20% time etc - as a subsidiary they have got out of the innovation business, at least any more than other established companies.
Alphabet makes more sense when you read it as Alpha Bet - http://www.investopedia.com/terms/a/alpha.asp
https://www.forbes.com/sites/miguelhelft/2016/05/19/google-c...
Source: I work here and take advantage of 20% time.
How did I not notice that earlier?
I makes perfect sense for Google to invest in robotics. They are already heavily invested in AI which will benefit robotics immensely. And with the coming robotics revolution, whatever companies get in early stand to make billions.
I was under the impression originally that they were going to use it for "internal space" mapping (maybe not big dog, but something more like atlas, or maybe a smaller quad). This was something that Google Glass was supposed to be for, but of course we know how people reacted to that.
It kinda made sense at the time - sorta. There are places which the mapping cars can't go, and the "human backpack" mappers are few (I would imagine - that gear doesn't look lightweight, so I would imagine you need to be somewhat fit to take on that job) - and still can't go certain places (and I doubt they would be allowed in certain areas). But if they could get systems "inside" that can map but are unobtrusive, they could offer that new set of data to advertisers as well.
For example - though a dying area - think about malls, and what advertisers (and the stores inside as well) could do with the information gained. It's one thing to go to google maps and get directions to a place, but it stops at the "border" (of say a mall). You still don't know where that store is inside (or outside) of the border. You have to consult a map, or whatnot. Maybe you part on the opposite side of where it is, and you have to walk further, or whatever. Being able to map that in a finer manner would be useful for consumers and the advertisers.
Apparently I was wrong in my assessment. I guess it's a good thing I wasn't investing on anything based on that!
Since there are not many businesses that have the return on investment of being #1 in online ads, they're forced to look for "moonshots". Not that they've found any that make money on a scale comparable to ads.
(The US military is all-Diesel. One fuel is used in everything from generators to aircraft. Gasoline tankers have no place on the modern battlefield, where there are no secure rear areas. The smallest standard military generator, a 2KW unit, weighs 138 pounds.[1] There's no suitable off the shelf powerplant.)
[1] http://www.deweyelectronics.com/assets/pdfs/spec-sheets/Rev-...
Deep inside you know that's what they want to build.
To be a bit hyperbolic: Aldebaran (renamed as Softbank Robotics) may not have struggled as much if they got bought by a better holding company.
Aldebaran was one of the first companies to commercially release a humanoid robot (NAO). I have one under my desk right now, it is an amazing combination of hardware and software engineering. You can program it using Python, C++ or a visual programming tool called Choregraphe. Novice users can write simple programs for the robot, however the robot is not very capable of doing more complicated tasks such as operating (physical) tools. It is however used as standard platform in the RoboCup (robot soccer) competition. The company is (was?) in the process of creating a human size robot called Romeo, which is being used by various research institutes in Europe and could be used for household tasks.
I wonder what would have happened if Google bought Aldebaran instead of Softbank!
Like you mentioned, Boston Dynamics and Aldebaran have different goals. If Google however is interested in making a household robot then I think Aldebaran would have been more suitable for them. Pepper and NAO have actually been used in public places like hotels and shops, but would you imagine one of Boston Dynamics robots being used there?
[0] https://youtu.be/fSQ-V8ipKn0?t=3m28s (in Dutch, sorry)
The challenges for Pepper focus more on Human-Robot interaction, which it is made for of course.
The Open Platform robots and HSRs have to do tasks like store groceries and set+clean a table, which require much more manipulation skills
This seems like a very interesting choice of name for a household robot... :-)
I don't doubt it was an excellent robotics platform that has survived among niche markets, but the critique of the company regarding Pepper as a product is definitely justified.
Would Google have even lanuched Pepper if they acquired the company? It seems that SB was in a rush to push out a humanoid product which probably gave them practical experience quickly with real use-cases. That is basically the opposite of Google's strategy where they only release sure-things with large markets. They don't seem to like experimenting in the marketplace the way Pepper seems to have done.
Fun fact and by coincidence, Andy Rubin's last official gig prior to this startup was head of Google's robotics division and oversaw Boston Dynamics.
Larry Page convinced Rubin to run the division after moving Sundar Pichai (Now Google's CEO) to takeover the Android project.
And so the tech world turns!
(1) https://en.wikipedia.org/wiki/Andy_Rubin
Edited per comment below for clarity
7.1
A novice asked the Master: "In the East, there is a great tree-structure that men call 'Corporate Headquarters'. It is bloated out of shape with vice presidents and accountants. It issues a multitude of memos, each saying 'Go Hence!' or 'Go Hither!' and nobody knows what is meant. Every year new names are put onto the branches, but all to no avail. How can such an unnatural entity exist?"
The Master replied: "You perceive this immense structure and are disturbed that it has no rational purpose. Can you not take amusement from its endless gyrations? Do you not enjoy the untroubled ease of programming beneath its sheltering branches? Why are you bothered by its uselessness?"
I never knew tank steering systems could be so interesting, though.
>The Robotics industry is more important in Japan than any other country in the world. Japan employs over a quarter of a million industrial robot workers. In the next 15 years, Japan estimates that number to jump to over one million and they expect revenue for robotics to be near $70 billion by 2025.[1]
https://en.wikipedia.org/wiki/Japanese_robotics#Japanese_rob...
When in Japan I've interacted with robots in a shopping mall: they had replaced their information stands by robots with ipads, so you could either ask them a question (didn't work due to the mall's noise) or just enter the relevant information on their belly's iPad-like devices.
'In 2006, AIBO was added into the Carnegie Mellon University Robot Hall of Fame with the description "the Sony AIBO represents the most sophisticated product ever offered in the consumer robot marketplace."'
'Rubin also went to SoftBank Group Corp. and negotiated a deal to include Essential in the firm's nearly $100 billion technology fund, but the investment was nixed after Apple joined the fund, according to a report in the Wall Street Journal. At the Code technology conference last week, Rubin confirmed the SoftBank talks and said the deal didn't happen because another investor that was a potential rival joined the fund and would have become an Essential backer. He didn't name Apple. "I just went to other investors," Rubin said. "I'm lucky to have good support from people that want to support these crazy, big, audacious things."'
And so the tech world turns :)
This hotel has been operating since 705 AD: https://en.m.wikipedia.org/wiki/Nishiyama_Onsen_Keiunkan
Planning for 300 years doesn't seem so insane in context.
Google wants to sell (now sold) Boston Dynamics, a robotics company they acquired in Dec 2013. People are speculating why, and some claim this is because of a "culture clash".
I no longer work for Google so what follows is pure speculation.
Boston Dynamics (BD) are very successful pioneers. But their algorithms are not based on Deep Learning (DL) principles. And Google is leading the world in Deep Learning and can apply it to anything they want, including robotics. DL based algorithms do not provide a complete robotics solution today but there is wide agreement that this is the best path forward for the field. Why is this important? The difference is robots that can walk vs robots that can dance ballet. The goal is "graceful locomotion" which will be an order of magnitude more adaptive, more energy efficient, and faster than the current generation of robots.
I am suspecting this is part of the culture clash. Imagine a meeting where Google comes to BD staff and says "we want you to toss out all the software you have written, retrain your engineering staff to use DL and other Holistic methods, and create a new generation of robots capable of learning rather than being programmed"... then I can see BD staff saying "Ahem. That'll take a few years". And Google might then say "Forget that".
Changing one person's mindset and stance from Reductionist to Holistic methods is a multi-month to multi-year effort; I have tried that and have rarely succeeded. Now imagine a company's worth of engineers that need to do that switch. It might be much better, cheaper, and faster to start a company from scratch and hire a more receptive crowd of engineers. And seed them with Googlers that already drank the Holistic Kool-Aid.
The hardware solutions BD created is know-how and intellectual property that Google currently owns since they own BD. When they sell off BD they could exclude the key patents or sell them as non-exclusive licenses. Which means Google could start a competing robotics company, use the BD patents and hardware know-how, and add a Deep Learning based software stack to run that hardware much faster than they could turn BD around.
PHB (2017): "I just went to a conference. From now on, everything has to use machine learning."
I wonder if they will anthropomorphize the speaker services like Amazon Echo. In the US these are bare bones listening and talking cyclinders. The Japanese might want to put an animated face on them for feedback. And make them mobile too.
https://techcrunch.com/2017/05/20/softbank-vision-fund-first...
Except for SCHAFT, who, apparently were also bought by SoftBank today so go figure.
A humanoid robot that can be used for military purposes will have a vastly greater number of civilian uses. I used to work as a welder in a shipyard, and believe me, there were plenty of times I could have used a robotic aid to assist a dangerous maneuver, where one side is reasonably safe, the other Guaranteed Instant Death if the safeties fail.
And if Uncle Sam wants to foot the bill for baseline research, why not? Pushing the coming of useful home/civilian robotics away because it could also be used for military purposes, to me, is akin to not wanting to smelt iron because you could use it to make a sword. In the meantime, people starve without metal plows, and the tech will be developed anyway, by someone else, who will do what they want.
(It is better to be the one deciding what to do with the tech than the one waiting pensively to see what will happen.)
He claimed to regret it in retrospect, only because he learned after the fact that the Nazis were further behind than he thought.
But the benefit of hindsight is not something we get beforehand. When he thought the risk real, he knew there was only one choice to be made.
https://en.wikipedia.org/wiki/Einstein%E2%80%93Szil%C3%A1rd_...
//
On top of all that, nuclear medicine has saved more lives than the bomb ever took, by multiple orders of magnitude.
//
Please don't get me wrong- I am not some war hawk. I do not want people hurt or killed, and I don't want to personally build any weapons. But the sort of hand-wringing that some people do over 'evil' technology... Knowledge doesn't work that way. Information is morally agnostic. And if it has the potential to be used in a harmful way, the best way to avoid it is to be the one that sets the leading tone. Anything else, IMO, is just wishful thinking.
My annoyance with Google's pearl-clutching has more to do with the fact that they treat the entire military need as 'bad', when in reality, 80%+ of .mil jobs aren't even combat-oriented, and the types of robots Boston Dynamics focuses on are clearly intended for pack mule functions or the like. Even if you wanted a strict no-weapon policy, there is a tremendous amount of work to be done in the armed forces.
Arguably the bomb itself has saved more lives than the bomb ever took, by forcing our leaders to resolve their differences diplomatically instead of starting another world war.
Of course, you can't really say in public that without sounding like a caricature from Dr. Strangelove.
But, survive them we did. As he reminded me, there were many people on all sides working as hard as they could behind the scenes to de-escalate tensions, smooth over diplomatic ruffles, calm the angry leaders. MAD was (and is) such a terrifying thought, and enough people understood the implications, that no button was ever pressed. And that is something to draw hope from.
Military robots aren't automatically bad. If used correctly they could stabilize the security situation in some of the most hellish places on Earth, ultimately facilitating massive humanitarian aid.
https://www.westfield.com/valleyfair/services/all-services/p...
Generally only small children play with it, and it has the affect of a children's companion.
Maybe I can just chalk it up to a compression error in the dictionary though. Perhaps fewer words is better.
Mitsubishi heavy industries would be the other prime candidate for aerospace...
Google, 2001: "Don't be evil."
Google, 2013: "We make military robots."
And now they're Alphabet, so the brand thing should've been ok.
[0] https://en.wikipedia.org/wiki/Don%27t_be_evil
[1] https://www.sec.gov/Archives/edgar/data/1288776/000119312504...
1. http://www.japantimes.co.jp/news/2017/03/16/business/tech/ja...
If you have to have someone driving the robot at a high level then it is less helpful in industry. The robot can't on command say "go pick up that box" or "go hang the drywall on those studs". That is an AI problem for open construction environments. For controlled environment industrial applications the dumb, non-dynamic repetitive systems work just fine.
Construction companies don't have that missing piece. SoftBank may have a shot at it.
What BD has done is indeed impressive. But the motors used by GR are simultaneously cheaper and (to my eye) more responsive. And it looks like they've work out a lot of the kinematics. They just need to commercialize the technology to make some big bucks.
And a big part of the reason the Ghost Minotaur looks more responsive is that it's smaller and so moves with smaller time constants and because it's executing movements programatically rather than through a feedback loop.
For each leg, you've got two motors per leg, and each motor just has two variables: angle and torque. Then you just have a 6-axis IMU for the chassis itself. All that tells you everything you need to know about the body and leg position, and the terrain you're standing on. I'd say that's way easier than a typical legged robot.
Consequently, it should be easier and cheaper to implement a given level of capability.
but they have plenty of things going on this is just another distraction from there core product groups.
"We make killer robots."
That's one way to square the circle.
Thoughts?
This was more than a random HN comment. There seemed to be several persuasive reasons why Google would do something like that.
Note that three years ago, the general sentiment was that Boston Dynamics was pretty scary: https://hn.algolia.com/?query=boston%20dynamics%20google&sor...
https://news.ycombinator.com/item?id=7103433
"Technically, since their acquisition Boston Dynamics, Google is indeed in the business of war robots: http://www.nbcnews.com/technology/google-buys-major-military... "
Is there a chance that Google stayed ahead of the curve and prevented Boston Dynamics from pursuing that?
It sounds like general fud against google but back then people had less to worry about.
Given the current political climate I would actually welcome a robot dystopia. as long as the robots were running it. not the politicians.
OK, but can I program the robots?
What if they were programmed to maximize our current pains?
It seems more likely that when Andy Rubin was placed in charge of Google's hardware, he made a string of robotics acquisitions in a long term bet on a household robot of some sort. In a short period of time, Google acquired 8 robotics companies, many of which specialized in various forms of locomotion which could be explored in parallel to find an optimal form of movement through unstructured environments.
These startups seemed to go nowhere since Rubin left not too long after the acquisitions were made. Hard to say why Rubin left at the time for sure.
So it seems more like Boston Dynamics were acquired for their incredible robotic locomotion, and Google's ambition to make a household robot (code named replicant, which is the term for an android in Blade Runner. By now I assume Rubin is a big fan), would fit with the fact that Boston dynamics adapted a lot of their previous efforts into designing lightweight low range robots suitable for navigating indoor unstructured environments.
Rubin worked hard on Android and was ready to release to public. Then Jobs called Google pals and they all got in one room. Jobs called him "fucking dickless asshole" and that he ripped off many features of not-yet-released iPhone. Rubin know it was bullshit, but Jobs demanded plenty of functions to be removed. To his suprise, Google bros did bow to Jobs.
Rubin never had the same level of respect for Page or Brin ever since.
More to read: http://www.businessinsider.com/steve-jobs-on-android-founder...
Except that the link you posted seems to disagree with your assessment:
"Ethan Beard, an early Android business development executive, told Vogelstein: "We knew that Apple was going to announce a phone. Everyone knew that. We just didn't think it would be that good."
Google had phone software that was ready to launch at the end of the year, but it looked awful relative to the iPhone, so it was all scrapped and delayed. "What we had looked so ... nineties," an Android engineer said."
Under the direction of Rubin, Google scrapped their independent phone effort and started from scratch to try to match the iPhone feature wise, even though Page and Schmidt had promised Jobs privately that they wouldn't copy the iPhone.
Fact is that Boston Dynamics does one thing, really, really well engineered dynamic locomotion. From what I understand there's very little ML/AI to it, just really good modeling and engineering.
However, no one has quite figured out exactly where this technology will be most useful or practical in the near future. Even (maybe especially) when they were owned by Alphabet.
as someone who did a bit of graduate research in that area, i would debate the "really, really well engineered" part of that statement. they brute-forced the engineering and developed energetically expensive machines that were great as demos but not practical as products. compare how long one of their robots last vs any legged creature of comparable size, and the poor energetics are obvious.
i would speculate that their company was a bet that battery packs would become light enough yet have a high enough energy density that it's robots could one day become viable products, but it doesn't seem like they're winning that bet.
that's in contrast to the segway, which was a bet that computing power was fast enough to dynamically adjust and balance human scale motion with high dynamic stability. they won the technical bet, but still lost the productization bet.
Irrelevant in a factory, where you can adapt things as much as you like - but crucial if you envisage people sending robots to environments they don't own - like a robot carer helping Grandma go to the store.
Needless to say, the profitable applications for this aren't especially short-term, so the ROI for this research is hard to predict.
If you prefer, you can substitute "Like a robot delivery driver, who must deliver to flats without lifts" or "Like a robot cop, chasing truant teens through a mall" [1]
They cost about $7/hr each to employ (~$62k annually since they're 24.7): https://www.digitaltrends.com/cool-tech/knightscope-robots-i...
That's one, and also a much smaller footprint than a wheeled vehicle.
> This how modern salad farming happens for instance.
And mushrooms too.
Mushrooms too.
A robotics mentor once told me "just use bigger wheels" for rough terrain rather than legs and fancy control algorithms. But stairs are particularly difficult for wheels. House bots will almost certainly need legs to maximize utility. Unless they are so cheap you can get one for every surface.
Tell that to these guys: https://www.youtube.com/watch?v=v1apqfrJYwk. I think one day we have have two wheeled robots with these kind of capability.
1. wheeled sidewalk robots. 2. wheeled robot with arms 3. walking robot.
Still incremental just have sellable robots the whole time.
There isn't any particular inefficiency in their walker balancing implementation.
but if we just look around, we find a huge array of locomoting, self-contained, relatively energy efficient beings to draw inspiration from.
for example, the tendons and muscles in your leg can return as much as 40% of the energy expended in locomoting. and your legs include passive mechanisms (like how your knee joint locks and the alignment of the cruciate ligaments inside it) to keep you mostly upright without a lot of active control.
nature has ingenious solutions to learn from and take advantage of.
The US Army was also trying some exoskeleton designs. The Raytheon XOS [2] was quite capable but needed an external power cable. The Lockheed HULC [3] finally reached a self-powered configuration, but battery life was too short.
None of these will be fielded. Maybe the next generation.
The Army's next big "mobility" buy is a replacement for the HUMMV, an armored truck from Oskosh. Smaller than the MRAP, but a comparable level of protection. Boring, but useful.
[1] http://www.military.com/daily-news/2015/12/22/marine-corps-s...
I imagine we'll see progress on this front when there's a breakthrough there. Barring that, you get some analog of the rocket equation: power source + fuel requires armor (for survivability) which means more power and fuel is required, etc etc
Another avenue to cheat: put extra armor just around the fuel/power source.
Wouldn't that make it heavier and take away the advantage of having an exoskeleton in the first place?
I imagined something like the Soviet "Shtrafbat", battalions made up of prisoners and other undesirables.
Of couse the shtrafniks will die in a terrible conflagration when they bump their battery, but think of all the expensive equipment taxpayers can buy for them, and a use for the prison population! Of course the prisoners will be rented to the state for a nominal fee, and there will be compensation to PrisonCo when their property is damaged. It's a win-win-win!
1) Establish consensus in favor of disrupting and retarding any movement toward war robots (or perhaps even coming to a more affirmative defense of the three laws)
2) Carry out the direct action associated with #1 above.
Andy Rubin and Larry Page were going to build "household helper" humanoid robots - no really, that was the grand vision - but it all fell apart.
They even have a cute (kawaii) mascot. So cyberpunk.