Artificial Intelligence-Empowered Automated Double Emulsion Droplet Library Generation

Double emulsions, with core-shell structures, are versatile materials used in diverse applications such as cell culture, drug delivery, and materials synthesis. A library of double emulsions with precisely controlled dimensions and properties would streamline the process of screening and optimization for specific applications. Microfluidic droplet generation offers precise control over droplet dimensions and properties, making it ideal for the preparation of droplet libraries; however, their preparation is tedious because fluid flow control and emulsion collection are typically performed manually and microfluidic devices are vulnerable to minor disturbances, requiring continuous intervention by skilled operators. To address these challenges, we present an artificial intelligence (AI)-empowered automated double emulsion droplet library generator. Leveraging a convolutional neural network (CNN)-based object detection model fine-tuned on a custom dataset, the system integrates decision-making and feedback control for automated droplet generation and collection. The system monitors droplet generation every 171 ms —faster than the reaction time of Formula 1 drivers —ensuring rapid response to disruptions and consistent production of single-core double emulsions. The library generator autonomously generates libraries consisting of 25 distinct monodisperse droplets with user-specified properties. This system significantly improves droplet-based experiments by reducing labor and waste, improving precision, and supporting rapid, reliable droplet library generation.