When I joined a course on microsensors given by Steve Senturia, Martin Schmidt, and Roger Howe at MIT in 1988, I saw the first video of the earliest rotating silicon micromotor. It was the beginning of a high-time of microelectromechanical systems (MEMS): everything we did was new [1], and the MEMS community was sure that we not only reached a new frontier (which might be true), but that overcoming this boundary would lead to solutions for the most pressing problems of human kind [2] (which might not be true). With the demonstration of a rotating micromachine as a key element in MEMS, we had the impression that basically all problems in our young field could be solved. In the past 20 years, we saw the explosive development of the field. We now have a good idea of the value of MEMS, which is in many respects different to what we anticipated 20 years ago, and which includes many new developments.