CorrAI: A Self-driving Lab (SDL) for the closed-loop optimization of corrosion protection technologies

The data, codebase and figure-generation scripts used by the paper CorrAI: An AI-Driven Automated Platform to Bayesian Optimize corrosion inhibitors in electrochemical systems are stored in this record.The contents of the package are: (i) the complete processed electrochemical data that was employed in the study, (ii) folders of detailed measurements of OCP, EIS, LPR, and chronoamperometry measurements associated with each data item, (iii) code used to accomplish Gaussian-process-based Bayesian Optimization of inhibitor compositions, (iv) code used to compute replicate statistics and regression measures, and inhibition-power reproducible to recreate Figures 1-9 of the study, and (v) all the code needed to generate figures necessary to reproduce The Data/ directory has the master dataset that was cleaned (Dataset.csv) and organized subfolders with both raw and processed files of electrochemical measurements. The Figures/ directory has the figure results utilized in the manuscript (heatmaps, correlation matrices, sampling maps, GP parity plots, and optimization trajectories). All the analyses, model predictions, and visualizations reported in the Results and Discussion sections are reproducible using the Python scripts that are included (optimization.py, Analysis.py, corrobotdataanalysis.py and db connection.py). This archive offers complete computational re-producibility of analysis and optimization bits of the CorRobot platform. Higher level code Hardware-control and automation code are excluded on purpose, because they depend on the system in ways that are not very general.