Putting Some Life Into the System

Abstract

In September 1986 a blue-ribbon panel of 52 experts met in Santa Clara, California to try to put a fence around the terra incognita on the map of modern control theory. For those concerned with the future direction of the field, it’s instructive to reflect on the following points noted in the panel’s final report [19]:

• These large-scale systems consist of many subsystems which have access to different information and are making their own local decisions, but must work together for the achievement of a common, system-wide goal

• … each subsystem may be operating with limited knowledge of the structure of the remainder of the system, which may be reconfiguring itself dynamically due to changing requirements, changes in the environment, or failure of components

• Some elements that appear in this research … which have not been fully accounted for in the past are as follows: … viewing reliability and survivability as key performance specifications, rather than issues of secondary concern.

• There is the need to include models of human decision-makers in designing distributed architectures.

• … controllers are required to be robust with respect to modeling uncertainty and to adapt to slow changes in the system dynamics … we lack at present a set of prescriptive methodologies that can be used to design fault-tolerant feedback control systems

• … control scientists and engineers must face head-on the challenge of increased complexity