Machine learning enables completely automatic tuning of a quantum device faster than human experts

An unavoidable obstacle to creating large circuits with spin qubits is device variability. Due to this variability, bringing a spin qubit into operation conditions requires a large parameter space to be explored. This process is becoming intractable for humans as the complexity of quantum circuits grows. We present a statistical algorithm that utilises machine learning to navigate the entire parameter space. We demonstrate fully automated tuning of a double quantum dot device in under 70 minutes, faster than human experts. This approach also provides a quantitative measurement of device variability, from one device to another and after a thermal cycle. This is a key demonstration of the use of machine learning techniques to explore and optimise the parameter space of quantum devices and overcome the challenge of device variability.