Artificial intelligence in science of measurements and the evolution of the measurements instruments: a perspective conception

Abstract

Measurement techniques always presuppose a theoretical model of the phenomenon under measurement. Usually the construction and the validation of the model underlining the measurement process is based on a priori knowledge. When the measurement process concerns complex phenomena, model analysis cannot be only based on a priori hypothesis, but it must be performed continuously. This intelligent activity can be included in the measuring system only if it is conceived and organized as an apparatus of intelligent perceptive systems. This new concept of complex measuring system leads us to consider it as a real metrological agency instance of the agencies already studied in artificial intelligence and robotics. The paper, starting from an analysis of the historical evolution of the concepts of intelligent instruments, proposes a scheme for complex measuring systems represented as multiagent metrological agencies. An example of metrological agency applied to the field of environmental monitoring is reported.