IDEAS Factory - Chemical Craftwork: Directed Assembly of Functional Patterns (Brianchell)

大自然是独一无二的，善于构建复杂的、多功能的、自我调节的系统，从细胞到我们。学习以类似的方式操纵复杂性应该可以为许多问题提供新的解决方案。在这个项目中，我们的目标是指导导电网络的组装，以便在产品的结构中编码信息（功能）。这与中枢神经系统存储信息和响应刺激的方式非常相似，我们的目标是物理实现一种称为神经网络的计算设备。为了实现功能，我们有两种方法：“学习”，其中网络可以被教导具有功能，以及“进化”，其中用于构建网络的参数作为基因，可以进化为具有所需功能的网络编码。很可能需要一个组合（进化以建立一个总体结构，并学习“微调”它），就像我们的中枢神经系统的发展一样，我们将使用导电聚合物，纳米粒子，纳米管和动态化学波来构建我们的网络，主要是电势来指导形成，电导特性来模拟它们。

Nature is uniquely good at constructing complex, versatile, self regulating systems from cells to us. Learning to manipulate complexity in a similar way should allow new solutions to many problems. In this project we are aiming to direct the assembly of conducting networks so that there is information (function) encoded in the structure of the product. This is very similar to the way in which the central nervous system both stores information and responds to stimuli, and our target is to physically realise a type of computational device known as a Neural net. To achieve function we have two approaches / 'learning' in which the net can be taught to have a function, and 'evolution' where the parameters used to construct the net act as a gene which can be evolved to code for nets with the required function. It is likely that a combination will be needed (evolution to establish a gross structure, and learning to 'fine tune' it) in much the same way as a our central nervous system develops.We will be using conducting polymers, nanoparticals, nanotubes, and dynamic chemical waves to construct our networks, mainly electrical potentials to direct formation, and conductance properties to characterise them.