On this DnA we talked about how builders and architects are using UAVs with cameras attached to scope out views. And that’s how the growing number of drones in civilian use in the US are mostly being applied — to film stuff. But under the radar there is a growing number of drone hobbyists who see unmanned aerial vehicles as robotic devices that can be customized for other uses.
Among them are
Jared Shier,
Tam Tran and Mindy King, left, architectural designers in the offices of Gensler, in downtown Los Angeles (above from left, Jared, Mindy and Tam, with their customized hexacopter as seen from Shier’s quadcopter.)
With the support of Gensler, they are trying to turn a drone into a flying 3d printer. Why?
DnA met with Jared (left of image, below) and Tam in a basement room of Gensler, watched test flights of a tiny “nanocopter” and a DJI Phantom 2 “quadcopter;” and saw a customized “hexacopter” print out a short spool of plastic thread. They explain their thinking behind the project and why they believe drones will one day be applied to automated manufacturing.
DnA: How did you get involved in this project?
Jared Shier: We started looking at the limitations of current 3D printing technology and realized that the biggest limitation has always been the build size. There’s always an x, y or z restriction or limitation and people seem to be just building larger and larger 3D printers rather than figuring out a way to completely free themselves of those limitations. Think of it as your printer at home, you’re limited to print whatever you have on an 8 ½” x 11” or 11” x 17” sheet of paper. We wanted to say, “well, we could print in an infinite distance in the x direction, the y direction, and then in height in the z direction.
DnA: And this took you to drones?
JS: Correct. Yeah, we researched hexabots and cars and all these different types of platforms, but it really just seemed like the hexacopter or any type of flying multirotor UAV was the route to go.
DnA: So the concept would be that you would attach a 3D printer to a drone and send it up in the air?
JS: On a very basic level that was what we were trying to do. We wanted to see if we could do it, and once we did it we wanted to analyze the limitations and try to figure out ways to improve it.
DnA: So if you were building say a room within a room where we’re standing, hypothetically you would have your drone up in the air spewing out melted plastic (and I’m simplifying things here). But it would be flying around making the walls; is that the idea?
JS: That’s the idea, but the melted plastic is something we always knew wasn’t going to be the build material that we were going to settle on. But what our main focus is right now is now researching alternate types of build materials that are going to have more realistic applications.
DnA: But the point is in terms of how the drone is involved; the drone is up there expelling the printable material? Or is the drone sending signals to another machine?
Tam Tran: You can think of it as your consumer black and white printer. If you remove everything except for the printer head, and you ask yourself, “How can I get this thing to print anywhere in this room?” And we looked at the robotic, platform-sized chair and we looked into UAVs and we basically attached the printer head to that drone. And then the struggle of the year was to try and get that drone to fly with this printer head printing something.
DnA: And did you accomplish it?
TT: We accomplished printing a single line extruded filament.
DnA: Was it worth the effort?
TT: Yes! It’s definitely worth the effort to see how far this technology can push itself.
DnA: But for both of you, the point of these drones is not to use it in the way that most of us understand these drones are currently used which is to film stuff. You’re applying it or you want to apply it to building something.
JS: Correct, we see this as a scalable prototype. This is something that could be scaled up to the size of a helicopter, or it could be used as a bunch of smaller litter hexacopters or octocopters, whatever it need be, flying around working together to build small structure be it walls, or disaster relief type shelters–
DnA: But what is the point? It’s a lot of investment and highly sophisticated soft and hardware to accomplish something like a wall, which people can build with their hands.
JS: Yes, people can build walls and you can send out workers, but what about places where you can’t send out a worker, or a construction truck or materials?
DnA: And what materials would you use because obviously the drone can’t be carrying a pile of bricks?
TT: That is what year two is for — researching a new material. We always saw the 3D printer as a year one proof of concept component, and year two we are studying different type of materials that can actually be more of a wall type, a thicker printing material. Right now a 3D printing material is only 2mm.
JS: If you wanted to think many years down the road, as Tam mentioned, let’s say there’s a disaster in Los Angeles where we have enormous beaches and lots of beach sand. This thing could go scoop up a whole bunch of beach sand and combine it with some sort of inorganic binder and using the beach sand as a aggregate can extrude that and build some sort of rudimentary shelters with it. That’s just one option that we’re exploring. And again, as you pointed out this thing couldn’t really carry a whole bunch of bricks and start laying bricks with mortar and everything. But we have also started looking at ways that these could be teamed together to accomplish a greater goal. For instance, you could look at mid-air refueling. That’s something that the military has been doing for decades now. So you could have this thing flying perpetually and you could have other types of designed robots that could then come in and swap out a battery pack so that this thing could keep flying or could drop off more materials so it could keep building, without end.
TT: And to add to that MIT has been the forefront of autonomous drones. They already have a drone that can take bricks and lay them on top of each other in a very designed pattern. The drone doesn’t construct a design itself; we design the structure and then the drone carries it out. It executes the design. And maybe down the line, we could be laying out the mortar and those types of drones would be putting the bricks on the mortar, eliminating construction workers.
DnA: That’s not great news for construction workers, is it?
JS: That was something I was going to make a comment to you about your question about what’s the point if people can build walls? People can build the structures, I mean look what’s happened with assembly lines. People used to be 100% of all assembly lines and now they’re being replaced more and more by robots. So there’s no reason why something similar couldn’t happen to the construction industry. And you could realize enormous cost savings from the developer’s size.
DnA: But the next stage might be that the robot could actually design buildings. What about when the robot could do the job of an architect?
JS (laughing): Well that’s something we’ll start lobbying against straight away.
TT: There’s a difference between intelligence and automation. We humans have the ability to design, but robots are really good at execution. We humans have a really hard time executing our designs because we’re limited by the natural forces. Like we can’t fly, we can’t do these things, but we have great ideas. Robots, on the other hand, we can send them out and do all these things and they’re not risking lives right? We’re risking dollars. And that’s the difference; our job is to design a place and we look as robots as our second hand to execute our design in a faster, quicker, more efficient, maybe even better way.
DnA: This has been a research project but it sounds like it’s also been a lot of fun. What have you enjoyed about experimenting with drones?
JS: It’s just really liberating to see the world from a different perspective. If you’ve ever wanted to fly or imagine what it’s like to be a bird this is the closest you could get to it without learn how to hang glide or skydiving and having that little moment of it. You can just see the world from different perspectives and gain a new appreciation on life for the built environment, it’s really incredible (below, Jared’s quadcopter flies over the Cornfields).
TT: I think the best moment when we were researching this project, when we were flying the sky out there those hexacopter and Jared had his obvious drone flying around us recording our research and there was a moment where he flew the drone to us and we watched the drone watching us. And it was such a meta moment of just “I’m watching the drone watching us!” And it was just a really spectacular thing.
Find more of KCRW’s drone coverage here.