Slowly but surely we are getting there. This made me think of when digital cameras first came to market. Nobody took them seriously at first.
I would not be surprised at all if within 10 or 20 years we have humanoid robots like these, but improved several orders of magnitude, available on the market.
It will be another technological revolution. If it can do chores I will certainly be the first on line to buy my first model.
That's a valid remark, although it is very general. For instance Geiger counters might not help with the actual problem of sorting in recycling centers.
I would be interested in the actual available sensors for discerning the materials in question, say between a glass bottle and a plastic bottle. Are there sensors that can discern between Polyethylene, Polypropylene etc. ?
Single stream recycling is largely solved. Unsurprisingly, humanoid form factors aren't optimized for conveyor flows of materials.
Lots of gates, air puffers, filters, size categorizers, etc. Maybe you could add an arm at the end, but there's no reason for the legs and torso of a humanoid robot.
Thank you for posting this video! Very interesting.
As for robots, I was envisioning a sort of multi-tentacle thing that hangs over the conveyor, with multiple cameras and manipulators that can quickly extend, grab a piece, and drop it into an appropriate bin. No need to have a humanoid form of course.
The humans who found jobs at those recycling centres might not cheer on you for automating away yet another job opportunity for them. Eventually there won't be any 'simple' jobs left while the supply of 'simple-job' people is unlikely to dwindle. What are they going to do then?
Installing drywall is unsurprisingly easy. Just that just needs a driver, a jack and a bucket of screws. And the sheets are large enough where you can cover a room with just a dozen. What eats up time is the taping and mudding. It's what sets an abject amateur from a pro.
My god when I saw the title I thought how awesome it will be to see mudding automated. I was disappoint.
I'm just starting on drywalling a repaired section of my basement. I'm only at the "get the drywall from the hardware store to the basement" phase, already that was hard enough and that was the easiest part. Ugh.
It also requires a lot of careful thought, observation, and planning if you want to get a flat wall that will be easy to tape, mud, and paint, and will look good when it's complete.
Like all skills, there's a big gap between "easy enough for an amateur to do an amateur job" and "results worth paying an experienced professional to do a professional job." Until we have real general AI, robots will not be capable of the latter outside of tightly constrained situations, like assembly lines.
Installing drywall in this sort of set-up is pretty easy. When you've got to cut holes for electrical or cabling, move around curves, do ceilings or use hoists to get stuff up... things become a lot more cumbersome and tedious.
Long way to go here; it's not a real world yet. For example, the robot didn't have to accommodate electrical switches or outlets. Also a lot of walls aren't multiples of the usual 4x8 drywall sheet. Often two full sheets plus some filler are needed to cover the entire wall.
The pros I know install drywall sheets horizontally, in part so filler pieces don't have to be done while squatting or bending. The filler is in the middle of the wall. I assume low back or knee pain isn't programmed in yet.
My ex-wife, for all her faults, that woman could tape and mud where you almost didn’t need to sand it. Sadly,
good drywall skills do not a marriage make.
It’s a cool demo. We’re definitely getting better at making humanoids. I would guess this is a pretty “hard coded” demo designed to work for this exact room setup only.
It’s going to be some time before we have the machine intelligence necessary to do enough for this robot to find a toolbox on its own, retrieve the necessary tools, unload the drywall from a truck, carry it to the room in question, install it, and complain about OSHA all autonomously.
Still, I’ve never seen this before. We are making progress. Just keep in mind that an actual robot that would do this commercially is probably 20 years away. Someone else said 10-20, and I’m inclined to think 20+ is more realistic. As in, you hire robots for your construction because it’s cheaper/better.
Im optimistic about 10 years... even if they are slow and clunky theyd be able to work 24/7 whereas human works cant or at least be optimized for tasks that are more difficult or dangerous for a human to perform. They might only take up 1% of the work force but i think that number will grow quickly
>It’s going to be some time before we have the machine intelligence necessary to do enough for this robot to find a toolbox on its own, retrieve the necessary tools, unload the drywall from a truck, carry it to the room in question, install it, and complain about OSHA all autonomously.
I don't think it's necessary to develop machine intelligence to do some of those tasks. The toolbox could could emit EM pulses so the robot can find it, the tools could be put in a specific order and put back the same way by the robot, the drywall could have easy-to-remove arrow stickers so the robot knows how to find it's borders, etc.
That actually brings up an interesting concept that never occurred to me.
We could get automation of complex tasks far earlier than I had thought. We're at this divide where we're getting capable with our understanding of motion and mechanics, but intelligently figuring things out is a massive roadblock.
What if, as you say, we make very dumb robots where their interaction with the world has been entirely laid out by humans. Ie, a humanoid robot making toast seems difficult. Yet if we give it the exact dimensions of the room, rf sensors for locations of shelving, cabinets, bread, etc, it seems possible with our current tech that the robot could be programmed to handle dynamic environments as long as the inputs were laid out ahead of time.
There's limited value in something like toast making, defining the inputs would be a hassle. Yet, for house building, or factory automation, if every piece, shape, location and dimension was known it seems a fleet of robots could build a house with quickness and ease, no "intelligence" needed.
Humans would need to lay a foundation to a spec of dimensions, but then the bots could go nuts. Neat!
What you're describing sounds like the way pre-fab homes are built. A computer model is created which specifies every part, and that gets sent to the factory floor where all of the wood is picked, cut, and assembled into pre-fabbed shipable components. As much work is done in the factory as possible while keeping all of the parts flat so they're easy to stack onto a truck. The cutting is all done by CNC machines, the assembly is all done using jigs and much of it could be done by assembly-line-style robots, moving stuff around can be done by simple robots as well.
The components get stacked on a truck by humans running machines, but the stacking order is pre-determined by the computer model, so this can potentially be automated by an oversized version of common manufacturing machines. The truck is driven to the job site by a human, but that might one day be done with an autonomous vehicle.
The final step is assembly on-site, which is done by humans running big machines. This is potentially the only part where humans need to be heavily involved, to make sure everything goes correctly together and to do the fine adjustments and small details.
The translated source document implies that they already mark the drywall with QR codes. Considering all the other markings I've seen on drywall, this should be trival to add at the factory.
Two and a half years ago I wrote a paper on "robotic automation" for a big financial services company. One of the conclusions was that, while the world of things was heretofore designed with humans in mind, in "the near future" (i.e. about where we are now), "stuff" would be designed for "machine-first" use, rather than "human-first" use.
This isn't surprising. We've seen this before in the transition from "animal engines" to "steam engines" (think - the whole infrastructure around the care and feeding of the animals etc transitioning to manufacture, fuelling and support for steam engines). We're starting to see it now in the (slow) transition from "carbon-fuel transport" to "electric transport".
> One of the conclusions was that, while the world of things was heretofore designed with humans in mind, in "the near future" (i.e. about where we are now), "stuff" would be designed for "machine-first" use, rather than "human-first" use.
This.
Just to provide an example, humanity already has self-driving cars for decades, and they are deployed and are extensively used in public transoortation. The trick to solve this problem was to not force human-designed solutions to an automation problem.
Thus, instead of trying to automate vehicles to run on roads, we have vehicles running on railway tracks.
The only reason we use drywall in the first place is because of the cost of labor + time. Plaster is far superior by almost any measure.
If you created some sort of plaster roomba/printer/extruder, you'd obviate the need for drywall. Could probably use manufactured or recycled lath of some sort.
Cost/sqft of Drywall is ~1/3rd that of plaster (1.50 vs 3.50-4.50). Most of the price difference is additional labor.
You still have to put up plaster board (basically drywall) first before you float the surface with plaster. I can't imagine plasterers are putting up plaster slats in stud walls anymore.
FWIW my house in Palo Alto (total remodel 2001-2003) is lath and plaster throughout (indoor and out), all done by union plasterers. They are a specialist sub niche in moderately high demand around here.
Weaker in what capacity? It's no weaker than drywall if it has lath behind it and much harder, though I will admit you need more (thicker) plaster to prevent cracks.
As someone renovating a 100+ year old townhouse with plaster and some newer drywall, this is sometimes true. Some of the plaster and lath (mostly ceilings) have cracks and it's a pain in the ass to fix when it's up on the ceiling.
But the main walls are freaking solid. Most of the walls and ceilings I'm needing to repair or replace are old drywall that was either not the proper type or just hasn't held up to wear and tear (and the neglect of the previous owner) over the previous 15+ years.
On the plus side, the drywall that needs to be repaired or replaced is easy enough for me to DIY but some of the really bad plaster/lath stuff is going to require a messy and laborious removal process.
Still, I love the rooms with plaster walls in good shape. Great sound insulation and looks awesome. It's a shame I can't afford to have more of the bad stuff replaced with new plaster.
Doesn't make much sense to me. Apart from the fact that a humanoid could walk to the workplace – which it probably can't. For the actual task it would be much easier to use a non-humanoid with a carriage and several axes.
Also it is a bit of a hoax because the whole humanoid reporting kind of suggests the robot is autonomous while one the image you can see a camera mounted on top of the wall and there's probably some huge computer in the back.
A benefit is having a multipurpose robot that can perform a variety of tasks that humans do using similar tools and techniques that we already use. I'm sure there's cases where that makes more sense than having more purpose-built machines.
> For the actual task it would be much easier to use a non-humanoid with a carriage and several axes.
True, but I think the idea is that you would be able to download the TileSetter update and use the same robot to finish the backsplash in the kitchen. Then, with the newly-released CoverWall module, you can put that robot to work gluing and applying wallpaper with perfect, no-seam precision.
A carriage and several axes probably wouldn't handle the manipulation of wall coverings that diverse.
Very nice. I'd like to see the video with no breaks or cuts, though.
It's interesting to see this done as a job for which the robot does not have enough manipulators. This would be easier with two hands to handle the board and one hand to apply the fasteners. On a production line, you'd have some way to keep the workpiece firmly in place while inserting fasteners. But they did it the hard way, with a humanoid form and only loose control over the workpiece. Which is what humans do. They often just push workpieces against a fixed object for guidance rather than go for full clamping.
So humans aren't the only ones that have problems with collated screw guns.
(You need to apply firm pressure until the screw is driven. Which might not happen if the bit is worn, or the screw in the plastic tape didn't get aligned correctly, or one of many other problems that can happen)
Ah - but can the robot exceed time and budget estimates, do a shoddy job and then weasel it's way out of warranty obligations? Jobs are safe, everybody - there will be a few more unions who will lobby for bans on robotic drywallers because of some made up safety reason.
49 comments
[ 1.8 ms ] story [ 87.5 ms ] threadI would not be surprised at all if within 10 or 20 years we have humanoid robots like these, but improved several orders of magnitude, available on the market.
It will be another technological revolution. If it can do chores I will certainly be the first on line to buy my first model.
I would be interested in the actual available sensors for discerning the materials in question, say between a glass bottle and a plastic bottle. Are there sensors that can discern between Polyethylene, Polypropylene etc. ?
Lots of gates, air puffers, filters, size categorizers, etc. Maybe you could add an arm at the end, but there's no reason for the legs and torso of a humanoid robot.
Example: https://www.youtube.com/watch?v=ODu2kbpVSXg
As for robots, I was envisioning a sort of multi-tentacle thing that hangs over the conveyor, with multiple cameras and manipulators that can quickly extend, grab a piece, and drop it into an appropriate bin. No need to have a humanoid form of course.
I'm just starting on drywalling a repaired section of my basement. I'm only at the "get the drywall from the hardware store to the basement" phase, already that was hard enough and that was the easiest part. Ugh.
Like all skills, there's a big gap between "easy enough for an amateur to do an amateur job" and "results worth paying an experienced professional to do a professional job." Until we have real general AI, robots will not be capable of the latter outside of tightly constrained situations, like assembly lines.
The pros I know install drywall sheets horizontally, in part so filler pieces don't have to be done while squatting or bending. The filler is in the middle of the wall. I assume low back or knee pain isn't programmed in yet.
It’s going to be some time before we have the machine intelligence necessary to do enough for this robot to find a toolbox on its own, retrieve the necessary tools, unload the drywall from a truck, carry it to the room in question, install it, and complain about OSHA all autonomously.
Still, I’ve never seen this before. We are making progress. Just keep in mind that an actual robot that would do this commercially is probably 20 years away. Someone else said 10-20, and I’m inclined to think 20+ is more realistic. As in, you hire robots for your construction because it’s cheaper/better.
I don't think it's necessary to develop machine intelligence to do some of those tasks. The toolbox could could emit EM pulses so the robot can find it, the tools could be put in a specific order and put back the same way by the robot, the drywall could have easy-to-remove arrow stickers so the robot knows how to find it's borders, etc.
We could get automation of complex tasks far earlier than I had thought. We're at this divide where we're getting capable with our understanding of motion and mechanics, but intelligently figuring things out is a massive roadblock.
What if, as you say, we make very dumb robots where their interaction with the world has been entirely laid out by humans. Ie, a humanoid robot making toast seems difficult. Yet if we give it the exact dimensions of the room, rf sensors for locations of shelving, cabinets, bread, etc, it seems possible with our current tech that the robot could be programmed to handle dynamic environments as long as the inputs were laid out ahead of time.
There's limited value in something like toast making, defining the inputs would be a hassle. Yet, for house building, or factory automation, if every piece, shape, location and dimension was known it seems a fleet of robots could build a house with quickness and ease, no "intelligence" needed.
Humans would need to lay a foundation to a spec of dimensions, but then the bots could go nuts. Neat!
The components get stacked on a truck by humans running machines, but the stacking order is pre-determined by the computer model, so this can potentially be automated by an oversized version of common manufacturing machines. The truck is driven to the job site by a human, but that might one day be done with an autonomous vehicle.
The final step is assembly on-site, which is done by humans running big machines. This is potentially the only part where humans need to be heavily involved, to make sure everything goes correctly together and to do the fine adjustments and small details.
This isn't surprising. We've seen this before in the transition from "animal engines" to "steam engines" (think - the whole infrastructure around the care and feeding of the animals etc transitioning to manufacture, fuelling and support for steam engines). We're starting to see it now in the (slow) transition from "carbon-fuel transport" to "electric transport".
This.
Just to provide an example, humanity already has self-driving cars for decades, and they are deployed and are extensively used in public transoortation. The trick to solve this problem was to not force human-designed solutions to an automation problem.
Thus, instead of trying to automate vehicles to run on roads, we have vehicles running on railway tracks.
Problem solved.
If you created some sort of plaster roomba/printer/extruder, you'd obviate the need for drywall. Could probably use manufactured or recycled lath of some sort.
Cost/sqft of Drywall is ~1/3rd that of plaster (1.50 vs 3.50-4.50). Most of the price difference is additional labor.
https://en.wikipedia.org/wiki/Lath_and_plaster
Drywall on other had is stronger by itself for the equivalent thickness, and MgO board beats any other wall panel on that.
But the main walls are freaking solid. Most of the walls and ceilings I'm needing to repair or replace are old drywall that was either not the proper type or just hasn't held up to wear and tear (and the neglect of the previous owner) over the previous 15+ years.
On the plus side, the drywall that needs to be repaired or replaced is easy enough for me to DIY but some of the really bad plaster/lath stuff is going to require a messy and laborious removal process.
Still, I love the rooms with plaster walls in good shape. Great sound insulation and looks awesome. It's a shame I can't afford to have more of the bad stuff replaced with new plaster.
Also it is a bit of a hoax because the whole humanoid reporting kind of suggests the robot is autonomous while one the image you can see a camera mounted on top of the wall and there's probably some huge computer in the back.
EDIT: Here's some actual information about the project: https://www.aist.go.jp/aist_j/press_release/pr2018/pr2018092...
True, but I think the idea is that you would be able to download the TileSetter update and use the same robot to finish the backsplash in the kitchen. Then, with the newly-released CoverWall module, you can put that robot to work gluing and applying wallpaper with perfect, no-seam precision.
A carriage and several axes probably wouldn't handle the manipulation of wall coverings that diverse.
It's interesting to see this done as a job for which the robot does not have enough manipulators. This would be easier with two hands to handle the board and one hand to apply the fasteners. On a production line, you'd have some way to keep the workpiece firmly in place while inserting fasteners. But they did it the hard way, with a humanoid form and only loose control over the workpiece. Which is what humans do. They often just push workpieces against a fixed object for guidance rather than go for full clamping.
https://www.bls.gov/opub/ted/2016/employment-by-industry-191...
(You need to apply firm pressure until the screw is driven. Which might not happen if the bit is worn, or the screw in the plastic tape didn't get aligned correctly, or one of many other problems that can happen)
Obesity will be an ongoing issue for humanity.