Exploratory tests for the design of a Python accelerator middle layer

Several laboratories and facilities recently started joined efforts towards the realization of a python accelerator middle layer (pyAML) for control, tuning and optimization. This software is intended as a successor to matlab middle layer (MML), inheriting its features but also expanding to new ones (e.g., nonlinear optics and machine learning tools). Presently, several codes are available that provide some of the desired features. These codes have been adapted and tested at several of the participating laboratories to give input to the design of the pyAML. The most relevant features and results have been analyzed and are presented here together with the implications for the pyAML design.