Robots is a website featuring regular podcasts about various aspects of robotics; currently it’s featuring a podcast of an interview with me by Sabine Hauert, from EPFL’s Laboratory of Intelligent Systems. This was prompted by my talk at the IEEE Congress on Evolutionary Computing, which essentially was about how to build a nanobot. Regular readers of this blog will not be surprised to hear that a strong theme of both interview and talk is the need to take inspiration from biology when designing “soft machines”, which need to be optimised for the special, and to us very unfamiliar, physics of the nanoworld, rather than using inappropriate design principles derived from macroscopic engineering. For more on this, the interested reader might like to take a look at my earlier essay, “Right and wrong lessons from biology”.
1 thought on “Soft machines and robots”
Comments are closed.
Interesting as usual. Good interview!
As I see it, one problem in building a self replicating machine is how to combine an external and internal coordinate system that in a logic way deals with discreet information (such as the instructions in DNA) – the machine needs to know “where to do what”. The coordinate system need to be able to copy itself during the replicating phase. I think the cell must use some sort of logical coordinate system – I have a problem understanding how the cell otherwise can coordinate it’s activities with such an extreme precision based on discreet information such as DNA.
I’ve tried to outline how to build a model inspired on the idea that the cell must use some sort of coordinate system where the logic by it self also represent a position system (both internal and external). The structure by itself would also be able to do at least do some computing. The coordinate systems 2D representation can also work as a replicating blueprint in building a 3D sphere which will get a surface that also can be used as a discreet coordinate system. The 3D sphere could then self replicate based on the 2D mother coordinate system.
I made a short simplified film showing my ideas. The starting point is next to trivial (even if I do think the “water calculator” can be useful in teaching binary for classes due to it’s simplicity) but the end is on the other hand perhaps a bit to speculative (but I wanted to provoke new ideas – being wrong can be a good way of doing so 🙂 ). The underlying math is extremely simple. I would very much appreciate your thoughts.