Recently, we successfully demonstrated in silico screening calculations of solid oxygen carrier particles for the CO2 capture technology Chemical-Looping Combustion (CLC). Previously, we have conceived, developed, manufactured, tested, and successfully demonstrated calcium manganate of single perovskite structure (Ca1-xMn1-yO3-δ) for this technology, which arguably constitutes state-of-the-art. In this project, we will develop these two promising concepts further. We will perform in silico screening of single perovskite materials for a related but more demanding application, namely the steam-iron process for hydrogen (H2) production with inherent CO2 sequestration. This will require a particulate solid redox material that not only has sufficient mechanical strength to be applied in fluidized-bed reactors, but also highly specific oxygen fugacity at relevant process conditions. This project involves applying an active learning cycle, consisting of an advanced generative model whose results are verified via Machine Learning Potentials (MLP) or Density Functional Theory (DFT) calculations, to identify feasible material compositions. The primary goal is to contribute to the development of new, more efficient, and more environmentally benign methods for the production of sustainable H2. The goal is to advance a material development procedure that can be used to develop other customized redox materials, for other processes with similar narrow demands with respect to thermochemical properties.