Fault diagnosis for flow-based microfluidic biochips

Advances in flow-based microfluidics allow biochemistry-on-a-chip for DNA sequencing, drug discovery, and point-of-care disease diagnosis. However, the adoption of flow-based biochips is hampered by defects that frequently occur in chips fabricated using soft lithography techniques. Fault diagnosis methods are now needed to improve fabrication processes and facilitate the (partial) use of chips that have defects. We present the first approach for the automated diagnosis of flow-based microfluidic biochips. The proposed method facilitates the identification of defects through syndrome analysis and a hitting-set problem formulation. The proposed technique is evaluated using three fabricated biochips, and exact defect localization and identification of the defect type is achieved in all cases.