Cognitive microsystems: geometry of computation and sensing

We describe one possible approach for the development of generation-after-next microsystems that have on-the-fly sensing and processing adaptivity. These microsystems should be viewed as the lowest level in a multi-tiered network. Ideas for the design, manufacture and integration of these microsystems borrow heavily from several evolving fields, including origami folding, computational geometry, topology, photonic band gap structures and three-dimensional nanofabrication techniques. This work is driven by the "geometry of computation and sensing". In other words, how do the spatial structures of computational and sensing devices define their properties, and, more importantly, can we use geometry as a design tool? Ultimately we envision families of microbots, built from cell-like modules analogous to stem cells, which cannot only learn and adjust to their environment but furthermore can adapt their form and function to accommodate possibly changing environments.