A new peer reviewed paper arising from collaboration under the ETERNAL RIA reports how a novel multi-agent Digital Twin approach has been applied to successfully simulate a microfluidic micromixer process for manufacturing DNA-Lipid Nanoparticles.

The new publication, entitled ‘Process Simulation of a Microfluidic Micromixer for Pharmaceutical Production of DNA-Lipid Nanoparticles’ appears in Processes 2026, 14(8), 1203, https://doi.org/10.3390/pr14081203 and is the latest result of a productive collaboration between ETERNAL researchers at IRIS Technology Solutions and Laboratorio Reig Jofre.

The research teams have successfully developed and evaluated a novel simulation of a highly industrially relevant advanced pharmaceutical manufacturing process (DNA encapsulation within lipid nanoparticles using a microfluidic micromixer). The simulation and the microfluidic approach which it models also address sustainability issues, such as reducing the environmental impact of the process itself, and reducing the need for physical testing.

The paper details the implementation and validation of the digital twin, taking into account key performance indicators and control parameters. The main method applied for development of the simulation is a novel multi-agent approach incorporating stochastic probabilistic behaviour, combined with theoretical definitions from the process experts at Reig Jofre and relevant literature, plus data and results from laboratory-scale experiments with different parameter configurations.

Implemented as a representation of the real physical process, the digital twin was found to reproduce the relationships between process parameters (flow rates) and experimental key performance indicators (capsule diameter, poly dispersion index, encapsulation efficiency) with general agreement between prediction and empirical results, providing useful predictive insights for laboratory experiments.

The simulation has potential as a support tool for laboratory experiments to reduce physical testing and indicate the most promising configurations on which to focus, with potential savings in time, resources, and other costs.

Read full open-access text* now

*Early access version published 8th April 2026. Complete PDF, HTML, and XML versions expected to be available soon.