Sensor Location in Distributed Parameter Systems

Many systems, such as flexible structures, fluid flow, geophysics, and climate, exhibit behavior that varies spatially in several dimensions, as well as in time. The issue in many applications is to deduce the state of these systems based on a mathematical model, as well as measurements made by sensors at a limited number of locations. A well-known example is weather prediction, which combines complex mathematical models with data obtained from measurements. Other examples are estimation of structural vibrations and the determination of the extent of oil deposits, contaminants and buried explosives. Similar issues occur in medical imaging. It is well known that the accuracy of the estimate is limited by the accuracy of the model and of the sensors.

The problem of determining sensor locations for distributed parameter systems is inherently interdisciplinary and requires the synergy of partial differential equation theory, numerical analysis and large-scale simulations, and inverse problems, as well as both frequentist and Bayesian inference and uncertainty quantification.

The aim of this workshop is to bring these disparate approaches together in the hope of achieving progress on the difficult but important questions related to sensor location.