The trouble with real numbers (Invited paper)
Comprehensive analytical modeling and simulation of cyber-physical systems is an integral part of the process that brings to life novel designs and products. But the effort needed to go from analytical models to running simulation code can impede or derail this process. Our thesis is that this process is amenable to automation, and that automating it will accelerate the pace of innovation. This paper reviews some basic concepts that we found interesting or thought provoking, and articulates some questions that may help prove or disprove this thesis. While based on ideas drawn from different disciplines outside programming languages, all these observations and questions pertain to how we need to reason and compute with real numbers.
Full Text: PDF