Microengineering: when is small too small? Nanoengineering: when is large too large?

Miniaturization, as exemplified by watches, was initially required for convenience. Cost-effectiveness was the drive for computer peripherals. New applications now push the research into microengineering and nanotechnologies. Mimicking animals has always been the dream of mobile robot builders. Nature has however developed extraordinary solutions for energy storage, scalable muscles with sizes down to few microns in diameter, multitudes of sensitive sensors and efficient distributed control mechanisms. Reducing the robot's size make them more appropriate to study collective behaviour, but at some point the intelligence must be reduced. Below one cubic centimeter, there is even no solution to just make the robot move autonomously. Nanotechnology open the door to smaller sensors and motors have been built in the 100 micron range. These motors are far from performing reliably, their control electronics are cumbersome and the assembly is critical. The artificial insect is not yet for tomorrow.