Assessment of microfluidic system testability using fault simulation and test metrics

In this paper we introduce a Microfluidic Fault Simulator, MFS, which uses a novel method of fault modeling and injection, the Fault Block, a generic and low abstraction fault modeling technique. This technique has been utilized over a wide range of fault conditions, in this paper we present a trapped bubble condition. In conjunction with injecting fault conditions, we can apply test methods. Two methods proving sensitive to microfluidic faults are; impedance spectroscopy and Levich electro-chemical sensors, illustrated here by a diffusional "Y" channel mixing system case study. Data from the MFS is analyzed using a Neyman-Pearson probabilistic approach, providing information on each sensor's test capability. Overall fault coverage for a given test is determined. This approach allows the analysis of fault coverage offered by functional-test orientated sensors to be compared to alternative approaches, which potentially offer increased coverage at lower cost