Machine learning based data driven inkjet printed electronics: jetting prediction for novel inks

Abstract

Machine learning (ML) as a predictive methodology can potentially reduce the configuration cost and workload of inkjet printing. Inkjet printing has many advantages for additive manufacturing and printed electronics including low cost, scalability, non-contact printing and on-demand customization. Inkjet generates droplets with a piezoelectric dispenser controlled through frequency, voltage pulse and timing parameters. A major challenge is the design of jettable inks and the rapid optimization of stable jetting conditions whilst preventing common problems (no ejection, perturbation, satellite drop, multiple drops, drop breaking, nozzle clogging). Material consuming trial and error experiments are replaced here with a machine learning based jetting window. A data set of machine and material properties is created from literature and experimental data. After exploratory data analysis and feature identification, various (linear and non-linear) regression models are compared in detail. The models are trained on 80% of the data and root mean square error (RMSE) is calculated on 20% test data. Simple polynomial relationships between the input and output features yield coarse prediction. Instead, small ensembles of decision trees (boosted decision trees and random forests) have improved predictive power for drop velocity and radius with RMSE of 0.39 m/s and 2.21 µm respectively. The mean absolute percentage error (MAPE) is 3.87%. The models are validated with experimentally collected data for a novel ink where no data points with this ink were included in the training set. Additionally, several classification algorithms are utilized to categorize ink and printer parameters by jetting regime (‘single drop’, ‘multiple drops’, ‘no ejection’). Categorization and regression models are combined to improve overall model prediction. Machine learning enables efficient material and printing parameter selection speeding up the development of novel ink materials for printed electronics by eliminating jetting experiments that are money, time and material intensive.

Machine learning enabled integrated formulation and process design framework for a pharmaceutical 3D printing platform

“…This can also increase the variance in volume of the primary drop and frequency of the secondary satellite drops. Accuracy reported for this model is comparable with values reported in the literature for other inkjet printer models that range from 70% to 90% 24,25,32,33 …”

“…Elbadawi et al explore multiple machine learning models for estimating the printability of formulations and the dissolution behavior of printed dosages for a wide range of formulation properties 24,32 . Brishty et al compare different machine learning techniques like neural networks, support vector machines, decision trees, and so on, for predicting drop size, velocity, and jetting regime for different materials and printing conditions 25 . Wang and Chiu use a data driven autoregressive model for estimating drop size for different printing conditions 33 .…”

“…In this study, the first part in developing the FPD framework is building a model to simulate DoD operation (predict product performance based on inputs). Data‐driven and machine learning models have been frequently used for this application as they can capture the complex fluid dynamics involved in inkjet drop formation with relative ease 24,25 . This study uses a category of machine learning techniques called artificial neural networks (ANNs) for modeling DoD operation.…”

Machine Learning Enables Process Optimization of Aerosol Jet 3D Printing Based on the Droplet Morphology

“…Therefore, it has the potential for efficient process optimization in various additive manufacturing (AM) systems. , Specifically, machine learning usually establishes cause–effect correlations between process parameters and the printing quality of AM in a data-driven manner to achieve effective process modeling and optimization. For example, based on different machine learning methods, the influence of main inkjet printing − /extrusion printing − process parameters on droplet/line behaviors was investigated and optimized, which will be beneficial to the wide application of inkjet printing/extrusion printing technology in the field of printed electronics and bioprinting. In recent years, some representative machine learning methods have been introduced into AJP to optimize printing quality.…”

Image Analysis Based Evaluation of Print Quality for Inkjet Printed Structures

“…An often-used method to influence wetting is to adjust the ink's properties, e.g., by adding solvents and surfactants and to determine their influence by contact angle measurements as conducted in [6,7] or [8]. To minimize testing efforts, [9] implemented a jetting prediction tool to improve jetting capabilities of newly formulated inks. Other approaches, such as [10], tried to improve the wetting by surface treatment with plasma or excimer laser treatment [11].…”