Just imagine what you could do with one of these. So bad ass.
I’m in the middle of hand building a sheet metal prototype with snips, brakes and rollers.
Would kill for this right now…
Holy awesome. I can’t wait for the trickle down consumer version.
Old tool, new spin. When will they be selling them at Staples?
I’m wondering how they’re dealing with the work-hardening issue. You’d think after the first couple of passes that things would start to distort.
yes, that was a pun 
Looks to me like they’re using a special alloy. Potentially aluminum.
A 5 axis CNC controlled English wheel is a pretty clever idea though.
I want to know how they calculate the tool paths. That is not a straightforward task at all.
Great tool, great concept.
I attended a Robots in Architecture workshop this week and saw part of a presentation where two “arm type” robots were forming sheet metal in the same manner from both sides. Robot programming, way simpler than I ever imagined. This workshop used a Grasshopper and Rhino plugin to program.
It seems like each pass works fresh metal in the examples provided. Maybe this issue is avoided if no multiple passes?
I was surprised at the simplicity of the path, like a perimeter milling pass. I am sure there is complexity involved with springback, but guessing that the placement of the backup tool sphere is a formula of the surface normal and unformed sheet direction.
Solving the work hardening issue…
I agree, double hexapod positioning, point form metal drawing, I can’t imagine more complex programming. The video shows software programmers have now displaced metal workers!
Someone correct me if I’m wrong but tool paths on this thing should be a breeze compared to high-res mills…it can’t even do undercuts so there’s not that much interference it has to consider; really only the maximum angle of the work piece.
Maybe it’s easier than I imagine, but you have to account for springback, and the fact that you are plastically deforming the material. The workpiece appears to be rigidly clamped, which would mean the thickness is changing as you shape it, in complex ways. With machining you are only ever subtracting the volume of the tool from the volume of the stock, that’s pretty straightforward.
With those robot heads, you could add undercuts. That would be very cool.
What if you did something similar, but using prepreg composite, and localized laser curing of the resin? Maybe I should patent that.
I suppose you’re right, though the angle of the work piece would have to get kind of insane, or maybe the tool heads’ range of motion is larger than I was picturing?