I think Robin Murphy, Texas A&M Professor, has the right suggestion for Google. No idea if Google will actually do it.
> If I were a large company with deep pockets, I would accept that robotics is what is called a ‘formative’ market and just like shopping on the web, it will take a decade or so to ramp up. Robots are innovations that do not directly replace people and thus it is hard for people to imagine how to use them—thus the market is formative. The theory of diffusion of innovation indicates that potential end-users, not developers, need to experiment with applications to see what works (and not just physically but human-robot interaction as well).
> However, robots have to be extremely reliable and the software customizable enough to allow the end-users to tinker and adapt the robots, so that the end-user can find the ‘killer app.’ Essentially, you can crowdsource the task of finding the best, most profitable uses of robots, but only if you have good enough robots that can be reconfigure easily and the software is open enough. I would concentrate on creating generic ground, aerial, and marine robots with customizable software and user interfaces in order to enable their regular employees (and customers) to figure out the best uses.
> In order to make sure the development was pushing towards the most reliable, reconfigurable, and open robots possible, I suggest the developers focus on the emergency response domain. Disasters are the most demanding application and come in so many variations that it is a constant test of technology and users. Imagine the wealth of ideas and feedback from fire rescue teams, the Coast Guard, and the American Red Cross, just to name a few! Focusing on emergency management would also have a positive societal benefit.
Xoogler here. There is zero chance that Google would get into this market. All the internal talk lately is about "only creating businesses at Google scale" which means $billions in potential revenue, or GTFO.
That is the only path not available to the dominating robotics countries of Germany, Japan and Korea who are literally under the gun not to develop such arms and weapons (without approval of the US).
Literally the US can dictate what kind arms they can develop. It is required they comply to receive consideration for US military aid and to remain under the US nuclear umbrella (from other nuclear armed countries like China, India, Russia, etc...).
Those countries also don't want to spend too much money on their military. So they can't develop nukes, their citizens don't want that. We actually want them to spend more, not less, like Japan. South Korea actually built automatic turret guns that shoot without human control. So I call BS on the evil us wants to stop them from building robots.
This if exactly the perspective at funding large projects at IBM Research. Interesting that the industry R&D labs fund more like top VCs in that regard.
That is not surprising since the robotics market is already dominated by other companies from Germany, Japan, and Korea. DARPA made the mistake of opening their competition to Korea, and Korea dominated the competition with an innovative robotic design. It was obviously far ahead in terms of practicality and their design could already be used today.
Google is trying to enter a market already dominated and pieced out.
As a German with an interest in computer and electrical engineering looking for a viable career path, I'd like to know, what companies are y'all talking about?
Agreed. The underlying assumption seems to be the employees at Google can spin up billion dollar businesses at will.
...which begs the question if I were able to spin up a billion dollar business at will, why would I work for salary and a minority equity stake at Google?
The emergency response domain is what you use when you've got nothing better in mind for an application of your hardware. It's a dead-end. (Except I guess for Murphy.)
I agree with the first 2 points, but the last doesn't follow. Disaster response isn't a good domain for end users to play around with different configurations to see what works. Few people are first responders, disasters rarely happen, and when they do it's no time to be experimenting with your homebrew robot.
Something that's closer to a consumer / small business task, and is tedious, common, and not critical to get right would be a better problem space to explore.
Isn't this an example of a Scientist speaking outside their area of expertise? Not substantially different from a Physicist speaking on Literature or History.
Personally, I don't see a mission-critical application for robotics any time in the immediate future. Low-risk deployments will yield data, not rescue teams. Firemen have no time to fool with experimental gadgets.
I really don't understand where GP's comment comes from...you don't even need to look her up; Robin Murphy's byline in the article itself makes it clear that she is a CS professor who directs a center for robotics. How could you possibly compare somebody with that set of credentials commenting on robotics to a "physicist speaking on literature or history"? As someone who's typically a bit skeptical of such allegations, the only explanation for GP's comment I can come up with is sexism... (But I'd be interested to hear any more charitable interpretations or an explanation from the comment's author -- I'm really pretty baffled by the comment.)
...what? She is a professor of CS & Engineering according to the byline in the article, and upon Googling her, I find no reference to her teaching economics or any other non-CS/Engineering discipline elsewhere. And unless ctrl+f is failing me, "econ" isn't mentioned anywhere in the article (except "economically" once in someone else's comment) or in this comment thread. So I have no idea what you're talking about.
She talks about econ and, as you diligently pointed out, that's not her area of expertise, exactly. OTOH I would assume that a professional does have a clue about adjacent topics.
> Firemen have no time to fool with experimental gadgets.
TI and Raytheon have used local police and fire departments as test customers for various electro-optical systems for decades. They are amenable, as long as they get the systems and support donated to the department.
Hardly any of the problems they suggest robots can solve are problems Google faces though. Last 10 metres delivery, supply chain control and management, disaster relief work, oil and gas, home assistants - particularly for seniors. There are too many comments in there to address all of them but here's a few:
"If anyone could crack the indoor social robot market that is seeing such high interest right now in both the consumer and commercial spaces, it would be them"
Why on earth would Google care about the indoor social robot market?
"If I were a large company with deep pockets, ... it will take a decade or so to ramp up"
It's always easy to say how other people should spend their money, and just because a company is one of the few that can do a thing, it doesn't follow that they should do it. A lot of the speculation seems to be of the "if I had a billion dollars, I would..." wish fulfilment sort.
I can see the point of self driving cars. They were doing street view anyway and that's a pretty obvious application for autonomous vehicles. But they're not actually a manufacturing or supply chain company. Their foray into that with Motorola was an expensive mistake. I just don't see that a lot of the other robotics stuff they're doing is relevant to their business. That doesn't mean nobody should be investing in this. Sure there are applications down the line, but are they relevant to Google?
The only one I can think of is data centre build out and maintenance. But Google aren't near the scale to support an entire industrial robot development and manufacturing industry just for a handful of data centres.
Having said that, are they profitable problems to solve? It doesn't really matter if they're things vital for google now, but if they're profitable and google are well placed to profit from them then it makes sense to do surely.
> Why on earth would Google care about the indoor social robot market?
Why on earth would google care about the home heating market?
I don't think the acquisition of Motorola was for anything other than the patent portfolio.
In https://www.google.com/press/motorola/ , Google mentions Motorola Mobility (the mobile phone arm of Motorola), but I believe this is overshadowed by the overt statement about patents immediately afterward.
They didn't really lose much on the motorola deals though. They got cash, partitioned the company out in chunks and got a bunch of patents. They probably got the patents cheaper than what they would have spent on just buying them.
Last 10 meters delivery is a super easy problem though. Put packages in locked boxes, put the locked boxes on an automated car, and instead of going to my house go to me. I'll use my phone to unlock the box, and i'll use my muscle to take the package.
The answer seems to be "not very much". Google bought all those robotics companies but didn't get them to work together. All the companies are still in their original locations. Nobody has a product. Even Bot and Dolly, which had a product, no longer seems to be selling it. Boston Dynamics is being sold because they don't play well with others.
The whole robotics exercise seems to have been a hobby of Andy Rubin, and when he left, nobody had a clue what to do. Google/Alphabet, through mismanagement, may have added negative value to the robotics industry. Google's secrecy here seems to be more about hiding management failure than protecting intellectual property.
Martin Buehler, the brains behind BigDog, is now at Disney, and I expect we'll see more mobile robots there. Disney has wanted this for years; around 2000, they hired Danny Hillis to work on theme park robots. They got some improvements to their anamatronics, and a dino robot that pulls a cart but gets its balance from the cart wheels.
The next killer app in robotics is probably really good bin-picking. Kiva's mobile platforms can bring the shelf to the picker, but a human still takes the thing out of one bin and puts it into another. Amazon is working on solutions to that.
After reading "Google parent Alphabet ushers in 'fiscal discipline era'"[1], it's likely that Google/Alphabet will effectively be out of robotics shortly, after trashing most of the R&D startups in the field. Nobody they acquired is 2 years from making big bucks. Workers will be, as previously threatened, reallocated to other parts of the business. Probably related to ads; that's still 94% of revenue.
This is discouraging. It's the acquire-trash-dump thing Paul Allen was notorious for.
> For autonomous deliveries, it's the last 10 meters that is the hard part.
Couldn't someone build a delivery box (like an american mail box) that could be mounted on your property, close to the curb, within reachable distance from a delivery truck.
The truck could stop, tell the 'amazon delivery box' to unlock and open. Then place items in it. There could be various different sized boxes, from small to large chest freezer type things. The box would have some electronics in it, a solar panel on top, it doesn't matter how expensive they are - you can rent them from Amazon or whoever.
This would then reduce all the complex variables of how to drop packages off at the destination.
This would be ridiculously useful for apartments, if you can find the space for it curbside. Large packages right now are delivered in one of two ways: 1) They leave it at the door. Risk of theft. 2) They leave it with the rent office, which is only open 9-5. Regular job? You have to either wait until Saturday, or you're screwed if your rent office isn't open Saturday.
I've thought in the past about trying to design a door door with a mail flap equivalent that works for really large packages, or something you could build into the wall next to a door to replace a coat closet that'd serve the same purpose. Moving it out to the curb and making it automation-capable sounds like an excellent proposal for this purpose.
You wouldn't even need one container per apartment/resident - just a reasonable number of various sized compartments. A keypad/pinpad or touch screen login would open the appropriate container for the customer. This is how Canada post handles parcel deliveries with their 'super mailboxes' (without the keypad part as they do have a small box for each unit where a key is dropped).
[edit] Authentication could be handled in a lot of different ways: QR code, RFID or other card/dongle/phone/internet based ways.
We have those in Germany, it's quite common. Opening is done with a card and code or barcode. The article mentions countries with similar systems: Austria, Finnland, Estonia, Latvia, Poland, Canada, Australia and United Arab Emirates.
edit: And not mentioned in the text but there is an image from Denmark.
In one of my previous houses (in Washington), we were on a hill with several other homes sharing the main private drive up the hill. We had a centralized apartment style mailbox with a slot for each house, and two big slots for larger packages. The postman had the same system of inserting the key for the bigger mailbox into your personal box.
20 years ago I was living in an apartment building where all the mailboxes were in a common location, with four large boxes for parcels. If you got a package, the mail deliverer would put it in a large box, and put a key to the large box in your personal mailbox.
In this age of amazon and online delivery in general, I don't think it's going to be that scalable, but it is an alternative.
Google bought Bufferbox a long time ago. They recently showed interest in actually having the autonomous vehicle have a bufferbox on it so you have to walk to the vehicle to get your package.
Google is doing research in AI to reach the singularity. I think robotic is very important for Google, but reassuring people is more important. Google does not want to be associated with something frightening like "terminator".
I think Google should continue with its self-driving effort. They have their maps and fiber technology that provide some hard-to-get pieces of the puzzle. This is the robot that will impact humanity the most. An automated box on wheels (seats optional). Just think of the things a self driving car could do for you:
- Pick up anything from mostly anywhere.
- Schedule pickup around the clock.
- Move things/people securely and be able to track the contents through a camera.
- Act as a mobile living space. Not requiring a dashboard frees up a lot of space.
That would mean that Google is able to fulfill a task with the robots, no? As a figure of speech, 'this should do' means that something is sufficient. Any first-language speakers?
It's not a reddit tracker, it doesn't get sent anywhere. It's added when you have the reddit companion extension installed, to tell it to show a toolbar.
It seemed fine by the time I saw it, but we took out the magic 17. The weirdest and most surprising of the tricks PG came up with for improving titles: https://news.ycombinator.com/newsguidelines.html.
> If I were a large company with deep pockets, I would accept that robotics is what is called a ‘formative’ market and just like shopping on the web, it will take a decade or so to ramp up. Robots are innovations that do not directly replace people and thus it is hard for people to imagine how to use them—thus the market is formative. The theory of diffusion of innovation indicates that potential end-users, not developers, need to experiment with applications to see what works (and not just physically but human-robot interaction as well).
> However, robots have to be extremely reliable and the software customizable enough to allow the end-users to tinker and adapt the robots, so that the end-user can find the ‘killer app.’ Essentially, you can crowdsource the task of finding the best, most profitable uses of robots, but only if you have good enough robots that can be reconfigure easily and the software is open enough. I would concentrate on creating generic ground, aerial, and marine robots with customizable software and user interfaces in order to enable their regular employees (and customers) to figure out the best uses.
> In order to make sure the development was pushing towards the most reliable, reconfigurable, and open robots possible, I suggest the developers focus on the emergency response domain. Disasters are the most demanding application and come in so many variations that it is a constant test of technology and users. Imagine the wealth of ideas and feedback from fire rescue teams, the Coast Guard, and the American Red Cross, just to name a few! Focusing on emergency management would also have a positive societal benefit.