DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact Dispensing
The reliable non-contact dispensing of droplets in the pico- to microliter range is a challenging task. The dispensed drop volume depends on various factors such as the rheological properties of the liquids, the actuation parameters, the geometry of the dispenser, and the ambient conditions. Convent...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Fluids |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2311-5521/8/6/183 |
| _version_ | 1797594821269913600 |
|---|---|
| author | Pranshul Sardana Mohammadreza Zolfaghari Guilherme Miotto Roland Zengerle Thomas Brox Peter Koltay Sabrina Kartmann |
| author_facet | Pranshul Sardana Mohammadreza Zolfaghari Guilherme Miotto Roland Zengerle Thomas Brox Peter Koltay Sabrina Kartmann |
| author_sort | Pranshul Sardana |
| collection | DOAJ |
| description | The reliable non-contact dispensing of droplets in the pico- to microliter range is a challenging task. The dispensed drop volume depends on various factors such as the rheological properties of the liquids, the actuation parameters, the geometry of the dispenser, and the ambient conditions. Conventionally, the rheological properties are characterized via a rheometer, but this adds a large liquid overhead. Fluids with different Ohnesorge number values produce different spatiotemporal motion patterns during dispensing. Once the Ohnesorge number is known, the ratio of viscosity and surface tension of the liquid can be known. However, there exists no mathematical formulation to extract the Ohnesorge number values from these motion patterns. Convolutional neural networks (CNNs) are great tools for extracting information from spatial and spatiotemporal data. The current study compares seven different CNN architectures to classify five liquids with different Ohnesorge numbers. Next, this work compares the results of various data cleaning conditions, sampling strategies, and the amount of data used for training. The best-performing model was based on the ECOmini-18 architecture. It reached a test accuracy of 94.2% after training on two acquisition batches (a total of 12,000 data points). |
| first_indexed | 2024-03-11T02:29:05Z |
| format | Article |
| id | doaj.art-b85f6315f2c84f289b6ab62f9f76ed5e |
| institution | Directory Open Access Journal |
| issn | 2311-5521 |
| language | English |
| last_indexed | 2024-03-11T02:29:05Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Fluids |
| spelling | doaj.art-b85f6315f2c84f289b6ab62f9f76ed5e2023-11-18T10:23:25ZengMDPI AGFluids2311-55212023-06-018618310.3390/fluids8060183DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact DispensingPranshul Sardana0Mohammadreza Zolfaghari1Guilherme Miotto2Roland Zengerle3Thomas Brox4Peter Koltay5Sabrina Kartmann6Laboratory for MEMS Applications, Department of Microsystems Engineering—IMTEK, University of Freiburg, Georges-Koehler-Allee 103, 79110 Freiburg, GermanyDepartment of Computer Science, University of Freiburg, Georges-Köhler-Allee 106, 79110 Freiburg, GermanyHahn-Schickard, Georges-Koehler-Allee 103, 79110 Freiburg, GermanyLaboratory for MEMS Applications, Department of Microsystems Engineering—IMTEK, University of Freiburg, Georges-Koehler-Allee 103, 79110 Freiburg, GermanyDepartment of Computer Science, University of Freiburg, Georges-Köhler-Allee 106, 79110 Freiburg, GermanyLaboratory for MEMS Applications, Department of Microsystems Engineering—IMTEK, University of Freiburg, Georges-Koehler-Allee 103, 79110 Freiburg, GermanyLaboratory for MEMS Applications, Department of Microsystems Engineering—IMTEK, University of Freiburg, Georges-Koehler-Allee 103, 79110 Freiburg, GermanyThe reliable non-contact dispensing of droplets in the pico- to microliter range is a challenging task. The dispensed drop volume depends on various factors such as the rheological properties of the liquids, the actuation parameters, the geometry of the dispenser, and the ambient conditions. Conventionally, the rheological properties are characterized via a rheometer, but this adds a large liquid overhead. Fluids with different Ohnesorge number values produce different spatiotemporal motion patterns during dispensing. Once the Ohnesorge number is known, the ratio of viscosity and surface tension of the liquid can be known. However, there exists no mathematical formulation to extract the Ohnesorge number values from these motion patterns. Convolutional neural networks (CNNs) are great tools for extracting information from spatial and spatiotemporal data. The current study compares seven different CNN architectures to classify five liquids with different Ohnesorge numbers. Next, this work compares the results of various data cleaning conditions, sampling strategies, and the amount of data used for training. The best-performing model was based on the ECOmini-18 architecture. It reached a test accuracy of 94.2% after training on two acquisition batches (a total of 12,000 data points).https://www.mdpi.com/2311-5521/8/6/183open microfluidicsnon-contact dispensingdrop-on-demanddeep learningvideo classification |
| spellingShingle | Pranshul Sardana Mohammadreza Zolfaghari Guilherme Miotto Roland Zengerle Thomas Brox Peter Koltay Sabrina Kartmann DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact Dispensing Fluids open microfluidics non-contact dispensing drop-on-demand deep learning video classification |
| title | DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact Dispensing |
| title_full | DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact Dispensing |
| title_fullStr | DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact Dispensing |
| title_full_unstemmed | DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact Dispensing |
| title_short | DropletAI: Deep Learning-Based Classification of Fluids with Different Ohnesorge Numbers during Non-Contact Dispensing |
| title_sort | dropletai deep learning based classification of fluids with different ohnesorge numbers during non contact dispensing |
| topic | open microfluidics non-contact dispensing drop-on-demand deep learning video classification |
| url | https://www.mdpi.com/2311-5521/8/6/183 |
| work_keys_str_mv | AT pranshulsardana dropletaideeplearningbasedclassificationoffluidswithdifferentohnesorgenumbersduringnoncontactdispensing AT mohammadrezazolfaghari dropletaideeplearningbasedclassificationoffluidswithdifferentohnesorgenumbersduringnoncontactdispensing AT guilhermemiotto dropletaideeplearningbasedclassificationoffluidswithdifferentohnesorgenumbersduringnoncontactdispensing AT rolandzengerle dropletaideeplearningbasedclassificationoffluidswithdifferentohnesorgenumbersduringnoncontactdispensing AT thomasbrox dropletaideeplearningbasedclassificationoffluidswithdifferentohnesorgenumbersduringnoncontactdispensing AT peterkoltay dropletaideeplearningbasedclassificationoffluidswithdifferentohnesorgenumbersduringnoncontactdispensing AT sabrinakartmann dropletaideeplearningbasedclassificationoffluidswithdifferentohnesorgenumbersduringnoncontactdispensing |