In engineering system control, human beings can play various key roles in particular concerning measurement, global assessment and decision. It is recognised that in such complex systems many involved variables are evaluated with uncertainty. In this paper, we used a possibility theory based approach to formalise all the different uncertain pieces of information. An applicative example concerning the safety assessment of dams is presented.

This paper compares two alternative averaging methods commonly adopted to estimate measurand values, in agreement with the guidelines of the Guide to the Expression of Uncertainty in Measurement (GUM). According to the proposed analysis, validated in a simple case study, the choice of best estimation method should depend not only on the amount of nonlinearity of the measurement model, but also on the amount of definitional and acquisition uncertainty of the input measurands.

Indications of the crisis in the conventional methods of metrological assurance of sensors are analyzed. The analogy between biological and technical evolutions is discussed. It is shown that the near future is mass production of intelligent sensors and systems with metrological selfcheck, as well as the transition to a new stage in evolution of metrological assurance of sensors.

The paper presents a fast impedance spectroscopy method for objects with very high impedance |Zx| = 1 GO modeled by multi-element two-terminal RC networks. The method is based on analysis of the object response in the time domain after square pulse excitation. The object impedance spectrum was obtained using continuous Fourier transform. Simulation tests were performed for different excitation pulse widths. Measurements in the realized system proved the usefulness of the method for parametric identification of equivalent circuits of the objects.

Aerodynamic balances are employed in wind tunnels to estimate the forces and moments acting on the model under test. This paper proposes a methodology for the assessment of uncertainty in the calibration of an internal multi-component aerodynamic balance. In order to obtain a suitable model to provide aerodynamic loads from the balance sensor responses, a calibration is performed prior to the tests by applying known weights to the balance. A multivariate polynomial fitting by the least squares method is used to interpolate the calibration data points. The uncertainties of both the applied loads and the readings of the sensors are considered in the regression. The data reduction includes the estimation of the calibration coefficients, the predicted values of the load components and their corresponding uncertainties, as well as the goodness of fit.