Fault modelling and test development for continuous flow microchemical sensor systems

In this paper we consider modelling and fault modelling of continuous flow microfluidic chemical analysis systems for the purposes of system design optimisation and implementation of test strategies. We demonstrate how a continuous flow system can be modelled using a static model which merges the chemical, fluidic and optical domain, and is used to implement a test strategy based on decision tree analysis. Decision trees are then used to highlight strategic test points within the system and where to implement self-test hardware. We demonstrate this approach with an example microanalytical system for measurement of glutathione in seawater using fluorescence. The test strategy employs impedance spectroscopy and we demonstrate experimental evidence of its basic feasibility.