A breakthrough in algorithms means each grain will know its place in a programmed shape.
This goes way beyond recycling. The ability to reshape things however you want, by reprogramming the stuff they’re made of, could revolutionise the efficiency with which we use materials.
So far, ‘smart sand’ is still a lab experiment. MIT Professor Daniela Rus and her team have created a working model of their new insights into granularity and structure. Complex as it may sound, their key breakthrough is developing algorithms that radically simplify what each grain in a box of ‘sand’ would need to ‘know’ to assemble into a given shape.
Telling each grain all the coordinates of its precise position would require ridiculous levels of computation. Instead, Rus’s smart grains just need to ‘talk’ to their neighbours. When a small template object is placed in the sandbox, those touching its outer surface map its outline – like starting a jigsaw with the edge pieces. Elsewhere in the sandbox, other grains can echo this shape. Those inside them can ‘know’ they’re filling in the structure – and others just fall away to the floor.
Ultimately, the idea goes, you could replicate any objects you put in your smart sand box – on pretty much any scale: keys or tools, for instance, of any size and number required... After you’d finished using them, you’d have the material re-sculpt itself to meet your next critical need.
Just think how this could cut the payload of a space mission. Backpackers might love it too – if they could ever afford it. Ultimately, it’s a concept that could change all sorts of equations in times of increasing material scarcity, in much the same way that 3D printing is revolutionising our idea of manufacture.
This brave new world is still some way off in practice – perhaps a decade even at astronauts’ prices, suggests Rus’s research student, Kyle Gilpin. The current experimental ‘grains’ are more like 1cm³ ‘pebbles’ – big enough to hold the electronics to define (and redefine) their roles, plus switchable on-off magnets on the sides so they stick to one another as required. And Rus and Gilpin have only really cracked the mechanics in 2D so far. But, they believe, they’ve seen the future – and it works. – Roger East
Photo: M. Scott Brauer / MIT