Synthesis of Reliable Digital Microfluidic Biochips using Monte

Synthesis of Reliable Digital Microfluidic Biochips using Monte Carlo Simulation Elena Maftei, Paul Pop, Florin Popenţiu Vlădicescu Technical University of Denmark

Background & Motivation Test tubes Automation Integration Miniaturization Robotics Microfluidics Slide soruce: Krish Chakrabarty, Duke University Automation Integration Miniaturization 2

Microfluidics Continuous-flow biochips (University of Michigan) 1998 Slide soruce: Krish Chakrabarty, Duke University Droplet-based biochips (Duke University) 2002 3

Biochip Architecture 4

Electrowetting on Dielectric 5

Synthesis: Main Design Tasks Allocation Binding Placement Scheduling 6

Fault Model § Types of faults: § Permanent § Parametric § Causes: § Particle contamination § Excessive voltage § Coating failure § Assumption: § Same probability for all cells 7

Fault Tolerance 8

Design Automation § ILP-based syntehsis § Constraints: § Scheduling and precedence § Resource § Placement § Optimization objective: § Minimize the completion time of the application § Evaluation of a design alternative using MCS § Generate a set of faulty cells § Try to reconfigure the biochip such that the execution time remains the same as for the case with no faults § Reconfiguration done using ILP-based synthesis 9

Experimental Evaluation § Real-life application § Mixing stage of polymerase chain reaction (PCR) § Assay completion time: 13 s § Experimental setup § ILP: GAMS 21. 5, CPLEX 9. 130 § Sun. Fire v 440 4, Ultra. SPARC IIIi CPUs, 1, 062 MHz and 8 GB RAM § MCS: 5, 000 runs with a cell failure probability of 0. 999 10

Contributions and Message § Contributions § ILP-based synthesis of digital microfluidic biochips § Considered permanent cell faults § Message § Labs-on-a-chip can be implemented with digital microfluidics § CAD tools are essential for the design of such biochips 11