The Art of the State [From the Editor

Abstract

Deep learning is impacting control in many ways, if not all deep. For more than a decade, it has reinstated the state-space concept as a central and unifying modeling principle. The magical equation <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x+ = f</i> ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x, u</i> ) currently dominates IEEE Control Systems Society publications, workshops, and conferences. Once confined to specialized circles of dynamic programming and nonlinear control, it has emerged over the last decade as a lingua franca shared by roboticists, optimizers, machine learners, computer programmers, and pretty much every scientist upgraded with the adjective “computational.” It celebrates the definite victory of computation over modeling in science and engineering.