It has long been thought that abiogenesis requires a process of selection and evolution at the molecular level, but this process is hard to explore experimentally. One solution could be the use of automation in experiments which could allow for traceability and the ability to explore a larger reaction space. We report a fully programmable and automated platform to explore the reactions of amino acids in the presence of mineral environments. The robotic system is based upon the Chemputer system which has well defined modules, software, and a chemical programming language to orchestrate the chemical processes including analysis. The reaction mixtures were analysed with tandem mass spectrometry and a peptide sequencing algorithm. Each experiment was screened for 1,398,100 possible unique sequences, and more than 550 specifically defined sequences were confirmed experimentally. This work aimed to develop a new understanding of selection in repeated cycles of polymerisation reactions to explore the emergence of well-defined amino acid sequences. We found that the outcome of oligomerisation was significantly influenced by the presence of different minerals, and that a serpentine environment selects glycine and phenylalanine rich fragments that enable the formation of longer oligomers with welldefined sequences as a function of cycle number.