Image Analysis, Automation, and Machine Learning Techniques Applied to MOS Quantum Dot Tune-Up

Tune-up and analysis of quantum dots (QD) is an arduous manual task consisting of a sequence of steps that builds upon one another. The tuning and analysis complexity is increasing as designs extend from QDs to multi-objects (e.g., donor-QD coupling and multi-QDs). The process can be simplified by utilizing image recognition techniques and automation. In this talk, I will present image analysis techniques which extract information from transport and charge sensing stability plots. These analysis modules can determine parameters such as tunnel rates and charge configurations in the QD systems. We identify the necessary combination of tune-up steps and feedback from analysis modules (i.e., output parameters for the next scan) that can automate tuning to few-electron charge sensing. This talk presents some of the proof-of-concepts, details and key future challenges.