In an era emphasizing high efficiency, AI technology is making a significant impact on the new energy industry. On June 18, ByteDance Seed announced an exciting collaboration with BYD to accelerate the development of lithium batteries. The partnership will leverage AI for Science combined with high-throughput experimentation to drive advancements in battery technology.

Partnership Announcement

On June 11, during the Volcano Engine Force Conference, ByteDance Seed and Volcano Engine revealed their deepened cooperation with BYD in the lithium battery sector. The partnership will involve the creation of joint laboratories and other collaborative efforts, focusing on leveraging AI to improve the performance of lithium batteries.

BYD’s Groundbreaking Megawatt Fast Charging Technology

Earlier this year, on March 17, BYD introduced its next-generation blade battery featuring "Megawatt Fast Charging" technology. This breakthrough technology has set global records, including a 50% reduction in internal resistance, a charging current of 1000A, and a charging rate of 10C.

With a voltage of 1000V and current of 1000A, the battery achieves the world’s largest charging power at 1 megawatt (1000kW), marking the advent of the "megawatt era" in charging power. This technology allows for ultra-fast charging, reaching a peak charging speed of "1 second for 2 kilometres" in mass production.

The Role of ByteDance Seed in Battery Development

ByteDance Seed’s innovative AI contributions are pivotal in BYD's success. During the early stages of development, BYD integrated ByteDance Seed’s BAMBOO AI model framework for electrolyte formulation. BAMBOO enabled rapid screening of electrolyte formulas, reducing the traditional trial-and-error process and accelerating research and development.

BAMBOO AI Model

The BAMBOO framework consists of two main components:

• ● BAMBOO-MLFF (Machine Learning Force Field): This machine learning-based force field is designed for liquid-phase organic molecules like electrolytes, combining molecular dynamics with AI model optimization.

• ● BAMBOO-Mix (Formula Property Prediction and Generation Model): This model is particularly useful during the broad screening phase, allowing for fast formula property predictions or generating potential formulas directly. It enhances efficiency by addressing issues like the time-consuming nature of molecular dynamics methods.

Future Collaboration and Focus Areas

Looking ahead, ByteDance Seed and BYD will deepen their collaboration through the establishment of an "AI + High-Throughput Joint Laboratory." This laboratory will focus on key challenges, such as fast charging, battery life, and safety. The partners aim to explore additional AI applications in battery R&D, accelerating the discovery of new materials and formulations, and further advancing power battery technologies.

Conclusion

By combining AI and high-throughput experimentation, ByteDance and BYD are pushing the boundaries of lithium battery technology, paving the way for faster, more efficient, and safer energy storage solutions in the future.