Traditional scientific discovery discipline always derives handcraft equations from first principles. Even with an expert’s careful mathematical modeling, questions always remain. With the rapid development of sensors, computational power, and data storage in the past decade, huge quantities of data can be easily collected and efficiently stored. Such a vast quantity of data offers new opportunities for the data-driven discovery of physical laws. From experiment design to data analysis, from physics law discovery to making useful predictions, my research aims to make the full process of scientific discovery fully automated.