Parchmint

Microfluidics based laboratory-on-a-chip (LoC) devices have been gaining traction in academia and industry in recent years. These devices are capable of automating biological experiments at mico-liter scale and below. The algorithms for design automation of these devices have been maturing in recent years yet there has been no development of a benchmark suite for use in analyzing the quality of algorithms from different institutions and research groups. We propose here a collection of real life LoC devices that have been specified in a standard notation, along with results from several recent placement and routing algorithms. This suite will enable researchers to quickly validate their algorithms and compare them against the cutting edge from other researchers.

Getting Started

Step 1:

Use the benchmark space images to identify the benchmarks you want to test your algorithm against.


Step 2:

Find the benchmark from our Github repository using the navigation bars links.


Step 3:

Compare the final JSON outputs against those placed and routed in the repository.

Step 1:

Fill out the fields necessary fields for generating the benchmark and click the generate button.


Step 2:

Download the benchmark using the download button.


Step 3::

Check to see if your algorithm is able to generate the layout for the benchmark.

Visualize

Upload the JSON file of the benchmark you wish to visualize. Clear the canvas when uploading new JSON files onto the visualizer.


Note: The visualizer only shows the benchmark as an abstract graph. It does not render the placed and routed benchmark that as it would have to be manufactured.

Standard




// Connection JSON

{
  "sinks": [
    {Target}
  ],
  "name": "String",
  "id": "String",
  "source": {Target},
  "params": {
    "start": [
      Integer,
      Integer
    ],
    "end": [
      Integer,
      Integer
    ],
    "key": value
  },
  "layer": "String"
}
              

// Target JSON
                            
{
  "component": "String",
  "port": "String"
}              
                                              

// Layer JSON
                            
{
    "name": "String",
    "id": "String"
}

Component JSON

The component JSON primarily contains the description of the component along with the information regarding the layers which contain the component, the parameters associated with the component and the complete set of ports which are available when routing the channels.

Port JSON

Each of the port objects contain the label by which the port can be identified, their relative positions on the component and the layer in which they belong.

Connection JSON

The connection JSON primarily holds the connectivity information and holds fields which store the source and sink infomation in the form of a target object. In addition to that it also stores the relevant parameters that might be used during the routing and fabrication stages.

Target JSON

The target JSON is a tuple that stores the id of the component and the port's label to which the channel connects to.

Params JSON

The params are ubiquitous key, value storage fields that are available for every object that is part of the standard. It allows for various tools to store technology specific information that can be utilized as seen fit by each individual tool. It also allows for new ideas to be tested and proliferated before being incorporated into the main standard.

Layer JSON

The layer JSON objects serve a link to group various components and can be used to easliy query the role of the component within the device.

Feature JSON

The features JSON is primarily used as a representation for visual editor tools. It allows for the separation of the objects that are necessary for automation from the object that would be used for visualing the micrlofluidic device. This part of the standard is currently under development is will be included in future versions of this standard. 


// Component JSON

{  
    "name": "String",
    "layers": ["String"],
    "id": "String",
    "params": {
    "position": [
        Integer,
        Integer
    ],
    "key" : value 
    },
    "ports": [{Port}],
    "entity": "String"
}
                          

// Port JSON

{
    "x": Integer,
    "y": Integer,
    "label": "String",
    "layer": "String"
}                
                

// Params JSON

{  
    "key1": value,
    "key2": value,
    ...
}

Generate

Scribe

Scribe is a tool that automatically generates benchmarks that can be used to test how Place and Route tools work against benchmarks of varying complexity.

Start Scribe

Publications

ParchMint: A Microfluidics Benchmark Suite 
B. Crites, R. Sanka, J. Lippai, J. McDaniel, P. Brisk and D. Densmore, "ParchMint: A Microfluidics Benchmark Suite," 2018 IEEE International Symposium on Workload Characterization (IISWC), Raleigh, NC, 2018, pp. 78-79, doi: 10.1109/IISWC.2018.8573487.
Specification, Integration, and Benchmarking of Continuous Flow Microfluidic Devices: Invited Paper 
R. Sanka, B. Crites, J. McDaniel, P. Brisk and D. Densmore, "Specification, Integration, and Benchmarking of Continuous Flow Microfluidic Devices: Invited Paper," 2019 IEEE/ACM International Conference on Computer-Aided Design (ICCAD), Westminster, CO, USA, 2019, pp. 1-8, doi: 10.1109/ICCAD45719.2019.8942171.