Seam carving is an algorithm that analyzes image content and can be used for size reduction in a manner that avoids direct compression or downscaling. Seam carving iteratively identifies horizontal and/or vertical paths of least visual importance and removes them from the image; each path removal reduces the length or width of the image by one row or column of pixels. This article adapts seam carving to reduce excess area of flow-based microfluidic chips that have been drawn by hand or by computer-aided heuristics without negatively impacting their functionality. The proposed approach leverages domain knowledge, wherein the image to be carved consists of I/O ports, components, and fluid channels, with known and understood fluidic behavior. Three different variants of seam carving are presented: 1) linear; 2) nonlinear; and 3) nonrectilinear; experimental results show that nonrectilinear, which is the most general of the three, yields the best results: it improves area utilization by 8.6× and reduces fluid routing channel length by 73% across a set of benchmark microfluidic designs.

Continuous-flow microfluidic devices based on integrated channel networks are becoming increasingly prevalent in research in the biological sciences. At present, these devices are physically laid out by hand by domain experts who understand both the underlying technology and the biological functions that will execute on fabricated devices. The lack of a design science that is specific to microfluidic technology creates a substantial barrier to entry. To address this concern, this article introduces Directed Placement, a physical design algorithm that leverages the natural “directedness” in most modern microfluidic designs: fluid enters at designated inputs, flows through a linear or tree-based network of channels and fluidic components, and exits the device at dedicated outputs. Directed placement creates physical layouts that share many principle similarities to those created by domain experts. Directed placement allows components to be placed closer to their neighbors compared to existing layout algorithms based on planar graph embedding or simulated annealing, leading to an average reduction in laid-out fluid channel length of 91% while improving area utilization by 8% on average. Directed placement is compatible with both passive and active microfluidic devices and is compatible with a variety of mainstream manufacturing technologies.

The discovery of conserved (repeated) patterns in time series is arguably the most important primitive in time series data mining. Called time series motifs, these primitive patterns are useful in their own right, and are also used as inputs into classification, clustering, segmentation, visualization, and anomaly detection algorithms. Recently the Matrix Profile has emerged as a promising representation to allow the efficient exact computation of the top-k motifs in a time series. State-of-the-art algorithms for computing the Matrix Profile are fast enough for many tasks. However, in a handful of domains, including astronomy and seismology, there is an insatiable appetite to consider ever larger datasets. In this work we show that with several novel insights we can push the motif discovery envelope using a novel scalable framework in conjunction with a deployment to commercial GPU clusters in the cloud. We demonstrate the utility of our ideas with detailed case studies in seismology, demonstrating that the efficiency of our algorithm allows us to exhaustively consider datasets that are currently only approximately searchable, allowing us to find subtle precursor earthquakes that had previously escaped attention, and other novel seismic regularities.

The lack of standardization in the specification and representation of microfluidic designs and their corresponding architectures is one of the largest hurdles faced by the developers of Microfluidic Design Automation (MDA) tools. In this paper, we introduce MINT, a Microfluidic Hardware Description Language (MHDL) for defining components and devices in a human readable manner, and ParchMint, an MDA interchange format and associated benchmark suite that can be used to compare the performance of different physical design algorithms. We further demonstrate how the introduction of MINT and ParchMint into the engineering workflow can bridge the gaps from the specification to the fabrication of microfluidic devices. While recent efforts to democratize microfluidics have been recognized by the community, there is an unfortunate lack of open source tools, design languages, and standards. Consequently, microfluidic designs shared on open platforms such as Metafluidics[15] leave conceptual gaps in terms of missing design information that are necessary to realize the “creative process flows” (Reproduce, Remix, and Test multiple systems). MINT and ParchMint are open source projects, which allows the community to contribute and extend their functionality to enable advanced algorithmic methodologies and new commercialization possibilities that differ from the vertically integrated industries we see today.

This paper introduces a technique based on seam carving to reduce the area of microfluidic very large scale integration (mVLSI) chips. Seam carving repeatedly identifies small slices of the device that can be safely removed (carved) and patched without adversely affecting device functionality. Using non-linear seam carving we achieve an average improvement of 4.28x in area utilization and an average reduction in fluid routing channel length of 53%

Continuous flow-based microfluidic devices have seen a huge increase in interest because of their ability to automate and miniaturize biochemistry and biological processes, as well as their promise of creating a programmable platform for chemical and biological experimentation. The major hurdle in the adoption of these types of devices is in the design, which is largely done by hand using tools such as AutoCAD or SolidWorks, which require immense domain knowledge and are hard to scale. This paper investigates the problem of automated physical design for continuous flow-based microfluidic very large scale integration (mVLSI) biochips, starting from a netlist specification of the flow layer. After an initial planar graph embedding, vertices in the netlist are expanded into two-dimensional components, followed by fluid channel routing. A new heuristic, DIagonal Component Expansion (DICE) is introduced for the component expansion step. Compared to a baseline expansion method, DICE improves area utilization by a factor of 8.90x and reduces average fluid routing channel length by 47.4%.

This paper describes an integrated design, verification, and simulation environment for programmable microfluidic devices called laboratories-on-chip (LoCs). Today's LoCs are architected and laid out by hand, which is time-consuming, tedious, and error-prone. To increase designer productivity, this paper introduces a Microfluidic Hardware Design Language (MHDL) for LoC specification, along with software tools to assist LoC designers verify the correctness of their specifications and estimate their performance.