Every smart sensor system relies on a thorough understanding of the underlying working principle of the utilized sensor. Here, a compact mathematical model is crucial to arrive at closed-form expressions that accurately describe the sensor behavior. While the system behavior in the linear region is typically relatively easy to obtain, especially the lower and upper end of the dynamic range, which are dominated by the system noise performance and its nonlinearity require a more sophisticated modeling approach.
At the Institute of Smart Sensors, we therefore not only deal with conventional linear time varying modeling approaches but also perform extensive research on the modeling of nonlinear dynamical systems in the absence as well as in the presence of noise. To this end, we have access to industry standard computer algebra software such as Wolfram Mathematica and Mathworks MATLAB.
While compact, closed-form system models are desirable to gain an initial insight in to the sensor system, system nonidealities, parasitic effects as well as nonsymmetric geometries often make it impossible to find closed-form models for real-world sensor systems. In these cases, numerical simulations can on the one hand help to investigate the expected sensor performance. On the other hand, especially in view of ever increasing computational resources, numerical simulations including parametric sweeps and optimization routines can be used to gain insight into the sensor system and improve the system performance.
At the Institute of Smart Sensors, we perform extensive numerical simulations on integrated circuits. To this end, we have access to state-of-the art software tools including the Cadence design suite and Keysight’s ADS and GoldenGate. Moreover, the Institute of Smart Sensors hosts the Department of Electrical Engineering’s competence center for multiphysics simulations. In this role, we not only teach multiphysics simulations using COMSOL Multiphysics to interested students but we also conduct research using a variety of industry standard software tools including COMSOL Multiphysics, CST Microwave Studio, Altair FEKO, MATLAB as well as custom FEM/BEM code.
While modeling and simulation are indispensable tools for designing a high-performance smart sensor system true evidence for an enhanced sensor performance or the proof-of-concept for an entirely new sensor system can only be produced by a prototype. Therefore, a large fraction of our research is devoted hardware realizations of smart sensor systems. Here, we frequently use the great potential of modern integrated circuit (IC) technologies to produce smart sensor systems with performances greatly exceeding the current state-of-the art. One of the specialties of our institute is the design of purely quantum or hybrid classical-quantum smart sensor system, where we try to push the limit of detection towards the quantum limit. Examples of such IC-based quantum smart sensors include our ESR-on-a-chip-based portable electron spin resonance (ESR) spectrometer and our NMR-on-chip technology.