Liquid Identification in a Microplate Format Based on Thermal and Electrical Sensor Data Fusion

Abstract

Multidevice synchronous measurement and thus a fusion of data from multiple sensor sources is becoming a fundamental but nontrivial task. Essentially, different sensors have their own limitations and uncertainties; the fusion of the data serves to increase the accuracy of performance. This research aims to expand the possibilities of multiparameter sensing by integrating impedance and thermal sensors into a microplate format for the simultaneous assessment of electrical and thermal properties. The sensor uses a modified transient plane source (TPS) measuring principle to monitor thermal changes and interdigitated electrodes (IDEs) for impedance analysis. The sensor is put to the test in the characterization of different glycerol and NaCl dilutions based on the varying thermal and electrical properties. At last, to demonstrate the added value of data fusion for liquid identification, a proof-of-application experiment shows the identification of fruit juices based on both measurands using a Gaussian mixture model. These experiments already highlight the power of this combined measuring technology in simple liquid identification processes, hence showing great potential to develop advanced diagnostics in various fields, such as healthcare, biology, or food quality by providing the ability to effectively perform parallel impedance and thermal-based measurements.