During the last thirty years microelectronics changed our lives and work significantly and consequently one should wonder about the potential of microelectromechanical systems (MEMS) that contain hundreds or even thousands of components. Would the impact of microprocessors with the ability of processing chemical or biological information within minutes or hours instead of weeks or years not be just as remarkable as the impact of modern electronic microprocessors processing electric information within milliseconds instead of weeks? [1] And would an enhancement of imaging array systems for media presentations to provide additional mechanical impressions not be a comparable step forward? However, the initial expectation thatmodified semiconductor fabrication technologies would be readily adapted to many other large-scale integration applications (LSI, 100 to 10 000 components) has only been fulfilled in very few cases such as the digital micromirror devices (DMD). [2] For most other microactuator applications the semiconductortechnology is too expensive and does not offer sufficientperformance and/or the necessary functionalities. Polymeric MEMS and their inexpensive fabrication techniques could offer a solution to this. [3] The first concept to fabricate and control
microfluidic LSI devices based on poly(dimethylsiloxane) (PDMS) was presented by Quake et al.[4] These devices consist of an integrated fluidic circuit controlled by several pneumatically driven control layers and are preferably used in high-throughput screening applications, simultaneously performing thousands of
identical operations. [5] Using pneumatic row and column multiplexers these devices can individually address a large number of elements, while a simultaneous control of several random elements is not possible. [6] Here, we present a further material-based approach of polymeric large-scale integrated MEMS based on a stimuli-responsive polymer, which itself acts as the active functional unit, i.e., the actuator. These devices provide a simple single-
layer set-up and allow simultaneous but individual control of a large number of random elements. Using an imaging array system, a so-called ‘‘artificial skin’’, as example we demonstrate the generation of visual and physical artificial impressions of a surface.