I'm going to be a pedant here - that does not look like a swarm, and they don't claim it is (in the video, at least). By "swarm" I would understand that all have the same programming and that the assembly process is emergent from simpler instructions plus exploration and feedback. Instead they appear to be programmed to act sequentially with a pre-planned set of actions. There is no feedback, no emergence. In short, it's nowhere near as cool as you seem to think it is. It's just a bunch of dumb machines doing exactly what they are told to do, with no smarts.
No smarts? From their behavior, I got the impression that the robots were assessing what was the next step in construction, which included assessing if another robot was still working on construction. They also had to avoid collisions.
Anything programmed by humans is a "dumb machine doing exactly what they are told." I think you meant that to mean each individual step was hardcoded. I don't think it was. I think they were all following the same algorithm. The group certainly spends a lot of time on path planning: http://repository.upenn.edu/grasp_papers/
Actually, you may be more right than you know (sorry). I've found this quote: "We tell the quadrotors what structure to build and they figure out the assembly plan and then build it."
https://www.youtube.com/watch?v=W18Z3UnnS_0&feature=play...
It still doesn't look very swarm-like to me (more like some kind of planner algorithm; I know the quote implies a distributed intelligence, but look at the way they queue always in exactly the same place, do laps, etc etc), but that certainly sounds more promising.
Correct about the swarm -- indeed it's not a multiagent problem -- but perhaps a bit unfair regarding the "dumb machine" part.
This appears to be a "multi-body" problem (rather than a "multi-agent") problem: that is, there is likely one primary decision-maker which dictates the construction task ordering, likely computed via a partial-order multi-body planner. That's part 1 of the AI.
But there's another part that's equally interesting. The multi-body planner produces a series of construction steps. Each step is ordered with regard to certain steps, unordered with regard to other steps (these can be done in any order but not simultaneously), and explicitly declared parallel with yet other steps (they can be done simultaneously). The quad-roters grab the steps and perform them in parallel when possible, and at least form a queue otherwise. Last, the quad-rotors have to perform these steps without hitting one another, so there's a significant real-time multi-body path-planning problem involved.
Multi-body planning, scheduling, and plan monitoring is nontrivial and should not be so lightly dismissed.
thank-you! (i'm sorry for "dismissing" the work; i do understand the amount of effort involved, but i was trying to emphasize what was not happening; thanks for the extra info on the unordered work, etc).
Hey, you can't have emergent properties before you have foundation properties. Can't build a building with magical emergent properties if you don't first figure out how to make a bot that can lift the beams.
This lab has several other videos with quad rotor robots like this (e.g. http://www.youtube.com/watch?v=MvRTALJp8DM), but don't worry about them escaping and starting a robot domination: they rely on infrared motion capture systems in that room for extremely precise location sensing.
Once GPS gets sub-meter accuracy, then we should start worrying.
Solutions based on expensive IMUs [0], SLAM [1], or visual odometry are more realistic than GPS. As a rule, never use GPS for fine-grained control. Even ignoring accuracy, GPS fails when you lack line-of-sight to GPS satellites, which is extremely dangerous on a quadrotor - how do you safely land a flying chainsaw with no feedback? When finely controlling a platform, you need accurate relative position. The global position is usually unimportant when the platform's position on Earth doesn't matter.
Sensors designed for computer vision applications are surprisingly light and small! Board-level cameras can fit in the palm of your hand, and reasonable IMUs are not much bigger than a matchbox. The biggest obstacles today are tradeoffs between performance/weight and performance/battery life. The platform needs algorithms with acceptable results, requiring almost no supporting hardware, and that are robust to the thousands of things that go wrong when trying to sense (noise, errors, resolution, occlusions, etc). Algorithms that fit all of these requirements are very uncommon. In fact, I would not be surprised if there was an Iron Triangle in there somewhere.
[0] Inertial Measurement Units: Devices that estimate acceleration (or some other measure of movement).
[1] Simultaneous Localization and Mapping. Using stereo vision, lidar, radio ranging, etc. to estimate your position, plus how the environment looks.
GPS + Accelerometer is already pretty darn accurate, I'm not really sure why they use the IR room. Perhaps simply easier to set up, and easier to co-ordinate the bots?
These things don't look very powerful, to put it mildly. GPS means outdoors which means wind which is a bit harder for a small, light hovering robot to handle. Indoor lets them demonstrate the principle with a useful simplifying assumption of no wind.
Yes, I did see the same videos. And I still think no wind is a useful simplifying assumption for them, particularly on power grounds.
A constant wind I agree looks like it would be within their capabilities, though their route planning software would need to take account of it and approach the target from upwind by default - not impossible but another thing to account for.
The real issue for me is gusting. They are definitely small and light enough that power-to-weight issues, particularly when carrying a load with a higher surface area to the wind than the craft itself, means that it will get knocked off course by gusts. That obviously needs allowing for when navigating to and from the construction site, but particularly when dropping off a load - if the release comes at the same point as the gust, the part simply won't end up where they wanted.
Do I think all of these problems are unsolvable? No, but it makes a useful simplifying assumption for their version 1 that they don't have to deal with any of them because they're flying indoors.
Depending on which waveband you use, different vision tasks become a lot easier. It looks to me like all the sensing is done by the room, not the quadrotors, hence why its all room-mounted.
Pretty neat! I'm sure they'll get to it eventually, but there's no advantage to having three drones in this demo, since the work isn't pipelined. It seems that a drone has to wait until one finishes picking stuff up before it can pick up something itself.
Interesting, but robotic cranes on the ground would seem to be more efficient and reliable in ordinary circumstances. It's not a question of how to grip it, It's a simple question of weight ratios.
What you say makes sense - robotic cranes should be able to operate with heavier pieces, and will require less energy per piece placement.
However, MAVs (micro air vehicles) could reach into places where cranes wouldn't be able - i.e. inside the building.
Also you can have many more MAVs working at the same time than cranes, so MAVs might be able to considerably speed up the building process.
There's no reason you can't combine these approaches. Using larger ground-based equipment to move girders and small, flying machines to do things like interior finishing makes a lot of sense.
A 4x8 sheet of 1/2" Gypsum board weighs nearly 60 pounds. A five gallon bucket of paint weighs about the same. A quadrotor couldn't finish an office because it would be too large to fit through the door...never mind the rotating blades.
You're right. But. As robotics becomes practical for building, we will not just replace human with robots. We will change how we build to make it easier to build with robots.
For instance, I was noticing that while you might not want to use magnets for everything in your building, you could use them as a guide during that intermediate period where the robot can't just put things together, but they can be mostly correct. Magnets could be used to do things like guide pieces of wood together with a sheath containing powerful magnets, then the wood can screwed together, then the sheaths recovered and used again on some other bits of wood.
The biggest challenge may very well be keeping codes up to date. I suspect that the crossover may very well be swift when it happens; robots are improving very rapidly and when they cross the line where they are a more cost-effective way to build a house, they aren't going to sort of edge up to it and pull alongside it, they're going to blow right past it and keep going.
Times are tough now but the next several decades still stand to be very interesting times, in all senses of the phrase.
(Oh, and personally I don't think flying robots have much future in construction. The economics are nonsense, excepting cargo helicopters, which are regularly used for some purposes today. I'm more speaking in general.)
It's funny, people have been talking about prefab houses (related, but not the same, as on-site robotic assembly) for years, but they still are a niche item. Despite all kinds of efforts to popularize them, such as:
I suspect that protectionism will delay the "robot revolution" long past the time when it's technologically feasible. Building codes will be changed in some way to require human labor, building material suppliers will be pressured to "run out of" the parts needed by robot builders, etc. It's been my experience that few people outside of tech actually like the idea of robots doing anything that a human could do.
If robotic ground based cranes can't reach inside the building, then for the most part, neither can human workers [making similar adjustments for scale]. Robotic helicopters might have advantages where helicopters are currently used for construction. But those are limited cases.
Generally, the duration of construction is determined by economics and weather. Flying robots probably don't provide an advantage on either front.
How about the combination for both. Perhaps there could be a 80/20 mix. It could very well make a lot more economic sense. It would seem like the cost vs moving weight ration would be much higher for MAVs compared to a crane.
A couple of years ago I've watched a documentary about crane operators working in a harbors. They were payed incredible salaries and it was stated that they won't be replaced by robots because of their capabilities in decision making.
Well, that and incredibly beneficial union rules. Lots of work printing cargo manifests, handing them to another person who types it into another system, and so on. But it's ok, that process only adds a marginal amount of delay, error, and cost. It's better than getting shut down by a strike, right?
I've been cleaning up after a flood disaster for the past few days, the most amazing technology i've seen in that time is the bobcats and their drivers. It takes 10-15 people a few hours to empty a house of all it's waterlogged belongings on to the street, and then 15 minutes for the bobcat to get it up on the back of a truck. When I see a tech demo like this I'm very optimistic about it's potential for disaster relief, where simple structures, shelters, and platforms would be of great benefit.
Great fun, but I suspect NIMBY concerns would kill any practical application.
Let's imagine they get the machines scaled to the point when they could build an actual habitable structure, and that they sort the power concerns that I suspect would make it insurmountably uneconomic.
Now, imagine a squad of sufficiently large and powerful helicopters buzzing away all day next to your office. How many people are going to be OK with that?
Why couldn't small drones build things out of small elements? A lot of our assumptions about what is a reasonable unit of work are based on examples with humans in the loop. It might make more sense to think of a fleet of these things as a form of 3D printer.
Lots of them working together means active traffic management becomes critical to prevent collisions. That'll be regulated because even if they're moving things the size of bricks rather than I beams, do you want to see what happens when two hovering robots each carrying a brick 5 metres off the ground bump into each other and suddenly stop hovering? Or if they start creating turbulence for each other and interfere with flight even without contact? Even if it doesn't get regulated by law the insurance implications would force de facto regulation.
Lots of small components means more joints. With a brick as our example that's fine but I'm not convinced the gain over a squad of fairly cheap bricklayers is there - with a girder frame structure as is more typical for large commercial premises that means for the same design it's simultaneously weaker, heavier and more expensive. Not a trio of attributes I'd want to apply to my structure.
This is a sort of technology I love, that's clearly very interesting and with all sorts of potential. But I don't see construction as the ideal target market for it, by a long way. Freight unloading would have been a great case had it not gone containerised some years ago. Warehouse order picking as with Amazon et al? Perhaps. You can have the storage floor as a controlled space without people underneath which reduces the safety issues and there wouldn't be the same problem with noise nuisance. If we wanted to go down that road though I'm not sure it's a big enough win over cranes / arms and conveyor belts, or that either is a win over fairly cheap people pushing trolleys (having worked alongside exactly that in the past).
If we want to stick with construction it would make good sense for use in hazardous environments because it would provide a way of deploying a large, highly mobile force to an area where you couldn't deploy human labour (or where that deployment becomes severely impractical / unattractive for whatever reason - see the film 'Moon' for example), but I'm not sure that's a large enough market to justify the expense of developing the tech enough to make it fully usable. There doesn't seem to currently be a shortage of labour prepared to go and work in Siberian gold mines or northern Canadian tar sands quarries, for the right price, and I can't see the power technology for these devices scaling to the point to make them likely cheaper any time soon to invest in it.
Cool tech searching for a better application IMHO. I'd love to see that better application though and would be heading down there to watch when it happened.
If it was next to my apartment building I'd want the drones to work during the day and if it was next to my office I'd want the drones to work at night. Most homes are not right next to office buildings. I wouldn't mind if the drones were a couple blocks away though.
I live next to a small industrial estate, my employers's campus location has houses the other side of the road and behind some trees from memory. Offices often have bars and restaurants around them which do good trade in the evening, sometimes city centre apartments mixed in too. I certainly follow the principle but I think there's less urban clear 'this time is OK' areas than you might think, low enough to further penalise this against traditional construction techniques.
Agreed, I'm just saying that between noise, safety and simple powering concerns, I think the window for this being superior to current methods is very, very small.
So, human workers and ordinary, loud construction equipment buzzing away all day for 2 years is ok, but robot equipment - quite capable of working 24/7 and likely capable of going faster - buzzing away for 8 months, thats unacceptable? I think unions are gonna scream louder than any NIMBY concerns.
24/7 robots are fine, but helicopters are _loud_, particularly those with the lifting power to do construction work. We're currently seeing small proof-of-concept machines lifting small, light parts, but the machines needed to lift useful construction parts will need to be many times larger and heavier, and if they're going to run for any serious amount of time (as would be needed to be useful) then they either need very much bigger batteries or large fuel tanks, which again means extra weight which means more power, more noise and more energy to lift the extra weight of the energy. It's not impossible but it's a vicious, not virtuous, circle.
I relatively regularly have the local police helicopter overhead and it's definitely loud enough to keep people awake, sometimes to wake you up in the first place. A whole fleet of these withing 1/4 mile of you will be loud, noticeably louder than a usual construction crew, and I'm still concerned about the neighbours complaining about the risk of this fleet of robot helicopters crashing into each other or dropping parts. Really, I think this is a very cool tech but a mad application of it that I really can't ever see working.
The shortsightedness of people worried about fewer jobs because of new technology is ridiculous. If you can build a house for the cost of land and materials only, the cost of living is lowered for everyone. With enough advancement and automation, the cost of life's necessities drops. As it decreases, the amount of human labor necessary to support day-to-day life will drop with it. Eventually we'll be able to support ourselves with very little laborious work, focusing instead on content production and other interesting and creative endeavors. Why try to prevent that?
Because although a grand vision of the future is wonderful, so is a guaranteed paycheck for the foreseeable future. For the people who'd be replaced by these machines, "eventually things will be amazing" is not much comfort in the face of "and you'll be replaced".
Robotic construction will eventually take over the construction industry. Especially once robotic transportation becomes possible.
There are many complex processes being performed at construction sites. Many steps take highly optimized machines to perform robotically but that a single human can do just by changing tools.
Because of these processes, automating the entire construction process would be very expensive to do right at the site.
All parts would have be modularized so they could be snapped together. Or robots will have to be able to change tools. Or parts will have to be moved from robot to robot, with as much work being done before parts are actually brought to the site.
Either way, it will be interesting to see how it works out.
This is awesome! It's sort of like a less precise large-scale 3D printer or MineCraft. I'm sure I'm not the only person who thinks this is certainly the beginning of automated construction. It works for assembly lines. We may finally be getting to the point where the same process will work in more mobile volatile environments.
The power limitation doesn't seem very difficult to surmount, a drone could just go swap its dead battery with one from a pool of batteries on chargers, and keep working.
They'd need two batteries in a drone to do that of course, or perhaps some other robot could execute the battery swap.
In another video they show 4 of them working together to lift a 1.2 kg structure, and they don't seem to have much trouble with it. It also shows a single copter lifting a piece of 2x4 several inches long, but doesn't give the weight. As just a ballpark guess I'd say these things probably max out at lifting and flying with a little under half a kg, they don't seem to have any trouble with at least .25 kg. The columns in the video linked at the top of the page are most likey a lot lighter than that, simply because there's no need for them to be heavy.
Inductive charger landing pads, a must have. Let them build forever. Just make sure the pads are plugged into a neighbors outlet ... or their Chevy Volt.
Great show, but there needs to be another device (carried by the quad rotors) to create very strong bonds between the parts. If the quad rotors could carry this bond-making robot that temporarily attaches to the parts and joins pieces with adhesive/nailgun/puzzle-piece fit, then they would be on to something.
Make a structure of actual use, like a tiny bridge across a moat that could hold people. The magnets have got to go.
What would be cooler would be a completely separate bond-building robot that can weld / whatever and can independently traverse the completed building. When it was too far from its next destination a quadrotor would come and give it a lift.
You might have several different types of crawler. A couple to receive materials from the quadrotor and hold them in place while a third follows a weld pattern.
Another approach would be to have a small electric lock built into the pieces. When a piece senses it is touching another piece this electric lock screws into place.
