Resumen:

Nanocomposite materials based on a polymer matrix and nanomaterials (as Graphene Nanoplateletes, 2D clays or Metal Organic Frameworks) acting as additives are an exciting type of materials that bring the opportunity to create plastics with better properties (mechanical, thermal, electrical, related with fire resistance, etc.) and more sustainable, specially if they are bio-based and bio-degradable. However, the amount of possible combinations of materials and processing parameters is over whelming and it requires a huge amount of experiments to find a composition with the desired properties.

To accelerate this process, we propose as the final objective of the thesis an autonomous laboratory for nanocomposite materials discovery, operated by robots and guided by Artificial Intelligence, and capable of preparation and characterization of the materials in a closed loop to optimize the composition according to the desired properties.

In the first iteration of the laboratory, developed during this first year of thesis, we employed 3D printing as the technology to produce specimens and focused on the optimization of 3D printing parameters using Bayesian Optimization. The workflow involves several pellet-based 3D printers working in parallel, a collaborative robot manipulator to handle the printed parts, camera vision to monitor different parts
of the workflow, and custom designed lab equipment based on Arduino with the objective of automating all the parts of the experiment. The results are incorporated in each step into a Bayesian optimization process, that analyzes the data and decides on the next experiments to perform. In this closed loop of optimization, we are able to optimize the density of the printed samples and test them for impact resistance and visual characteristics. The number of experiments needed is reduced and the human intervention is limited to maintenance tasks.