Machine Metabolism: Structure-Reconfiguring Robots

truss reconfiguring robot

It might be difficult to tell from the picture, but you’re looking at a robot that is capable of building and disassembling simple truss structures. We’ll let that sink in for a moment.

[Jeremy Blum] finished his metabolic machine research back in 2011, but just this month has had his journal paper published in the IEEE Robotics and Automation Magazine on Structure-Reconfiguring Robots.

The concept behind this robot is biological metabolism – the ability to break down nutrients into building blocks, and then to use them to build new things. What if we could build a robot to emulate this most basic aspect of biology? Well, they have. Take a moment to imagine the implications in space: a fully automated deployment (or repair) of large structures. Or back on earth, large radio towers that are automatically assembled, welded, and even repaired if need be. The possibilities are amazing.

To see the Structure-Reconfiguring Robot in action and to learn a bit more about how it works, check out the video after the break.

18 thoughts on “Machine Metabolism: Structure-Reconfiguring Robots

  1. So lets combine a bunch of ideas.
    Let’s assume we can build a similar robot that can carry loads / grab I-beams, and do welding.

    Now combine that with 3d building printing:


    Factory build skyscraper kits:

    You can 3d print foundations for a building overnight.
    All horizontal parts of the building are pre-made in a factory by humans and/or robots (including the floor carpets and tiles, the ceiling, lights, horizontal electrics and pipes).
    The kit includes the vertical parts of the building in different stages of assembly, laid out on the horizontal parts of the building, waiting to be put right and assembled.

    The entire kit is delivered on-site, and dropped/slotted into place by the same pick and place system used for the 3d printing.

    Smaller “structure-reconfiguring” robots grab the dis-assembled components for the “vertical” parts of the building, and put them upright. They slot them vertically and screw/weld them into place.

    They can then climb the vertical parts of the building, bringing with them the next level of the building, including floors or base for the 3d printer.

    Repeat until done.

    Same concept could also be applied from the “top-down”: Once you build your foundations, you drop the TOP floor kit first, build it, and use the robots to lift it by one floor. Then slot in the ” ‘previouse-floor-level’ -1 ” floor kit and build it underneath the previous one.

    Repeat until building.

    Since everything is done in CAD, if you push this to it’s logical conclusion, you could design a building to client(s) spec, have almost all the components/kits built in parallel at different factories while you wait for the 3d cement to set. Have everything arrive in a “just-in-time” fashion, with a minimum amount of workers to monitor the robot’s progress.

    Robots don’t sleep. So that’s 24/7 construction. All the hard parts are done in factory. All the easy parts are done on-site by robots.

    Future’s looking fun.

    1. Yes, thats really fun. And BTW if to use “non-Zen” variant with some extra frames as scaffolding, the entire construction of large and complex objects can be even more easy. Robots will use it as a “3D-railways” with exact positioning. And after constructing is finished they will just automatically disassemble scaffolding.

      1. Structures fit pretty well here in Australia. Pretty well meaning 10’s of millimetres over 100’s of metres.

        If you’ve got a robot assembling a structure, it’s not unreasonable to assume you’d have a robot manufacturing the structural pieces.

    1. Saw that, was amazing. Cells contain tiny 3D railways, and little engines crawl back and forth across them delivering chemicals to where they’re needed. Then the rails are dissassembled and rebuilt somewhere else as needed. All this happens within the previously thought of as squishy thing that is a cell. In such a ludicrously quick timescale that the rise and fall of intracellular empires looks squishy to us.

      It’s amazing how big we are. Every cell contains thousands of little machines, all working precisely, except when they don’t, in which case previously-specified measures are taken. In a tiny clockwork symphony that makes Metropolis look like a singing tea-kettle. The technical manual for a human cell would be gigantic. And then on a larger scale there’s bloody billions of them, making complex machines that the cells had no plans for, and were never written down.

      Chaos theory’s impressive when you see psychedelic pyramids on your screen, but when it comes to just BEING, it’s trancendentally bloody amazing!

      It’s amazing how slow we are too. The events that make us are sparks in nothing. I suppose that’s why people would blame life’s existence on religion. It’s gigantic, it’s massively sophisticated. But each system makes sense at it’s own scale.

  2. It needs to be able to pass a truss from one end to the other. Or have some other method of carrying trusses that doesn’t impede it’s ability to move.
    It can’t, for instance, remove a truss from the bottom of a U shape and install it at the top.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.