Lithium ion batteries and solid oxide fuel cells have become popular as clean energy devices. However, the commercial application of the batteries as complex electric power systems requires the batteries with long-term, multi-dimensional and high-precision performance. Some battery performance prediction methods are still in the initial stage of exploration. With the popularization and promotion of artificial intelligence, machine learning based on the artificial neural network technology has attracted recent attention. Recent advances in data science, such as machine learning, provide the scientific and engineering communities with flexible and rapid prediction frameworks, indicating great application prospects in materials research and development. This review summarized the latest advances on machine learning for the development of renewable energy technologies (i.e., solid oxide fuel cells, lithium batteries) and carbon reduction technologies, and gave some comments for the future development directions.