Millions of possible crystalline nanoporous materialsa,b have been identified for carbon capture, extending far beyond our capability to quantify the in silico adsorption performance of each individual nanopore by brute force calculations. Experimentally fabricating and measuring the adsorption properties of each framework is also unrealistic. A pre-screening step is required for better resource allocation. In this talk, we present our work which optimises the classification mechanisms for characterizing nanopore structures, enabling efficient high throughput nanopore screening. Our cloud-based automated nanopore screening tool applies computational geometry and topology to compute topological representations that encode scientific information for each nanoporous structure. Next, Grand Canonical Monte Carlo (GCMC) simulations are performed to calculate target adsorption figures of merit. These can be combined with machine learning to help in accelerating the estimation of adsorption properties based solely on the atomistic structure of materials. a) https://doi.org/10.1038/natrevmats.2017.37 b) https://arxiv.org/abs/2009.12289 *Tom Peters acknowledges funding from IBM Research, under its OCR and SUR programs in award IBM-TJP-6328340.