AI-Driven Energy Transition: Expanding Roles for Batteries [Energy-topia]

Since March 11, "InterBattery 2026," the largest battery exhibition in Korea, has been taking place at COEX in Seoul. If there is one overarching theme running through this year's event, it is artificial intelligence (AI). AI is bringing about sweeping changes in the battery industry as well.

The most essential requirement for expanding AI is power infrastructure. Major U.S. big tech companies such as Microsoft (MS), Google, Meta, and Amazon are prioritizing renewable energy sources like solar and wind power when building AI data centers. In 2024, MS signed a direct purchase agreement (PPA) for 10.5 GW of renewable energy with Brookfield.

To ensure the stable supply of highly volatile renewable energy, energy storage systems (ESS) are indispensable. This is the reason why demand for ESS has surged, particularly in North America.

During his keynote speech at this exhibition, Je Young Kim, Chief Technology Officer (CTO) of LG Energy Solution, introduced a forecast from SNE Research stating that the battery energy storage system (BESS) market size will triple by 2035 compared to 2025, adding, "Personally, I believe the market will grow even more than that."

At this year’s exhibition, all three major Korean battery companies—LG Energy Solution, Samsung SDI, and SK On—placed ESS at the center of their booths. As the electric vehicle market has experienced an unexpected slowdown in growth, K-battery companies are expecting ESS to become a new breakthrough.

AI needs more than just ESS. AI data centers must protect valuable data even in sudden power outages. This is where uninterruptible power supplies (UPS) or battery backup units (BBU) come into play. In the past, UPS relied on lead-acid batteries or diesel generators, but recently, they are increasingly being replaced by lithium-ion batteries, which offer superior instantaneous response and output characteristics. UPS units are installed in separate battery rooms, apart from server rooms. BBUs are designed as integrated units within the server racks themselves.

Companies are responding to this market with batteries tailored for AI data centers. Samsung SDI plans to launch a 2170 (21mm diameter, 70mm height) cylindrical battery specialized for BBUs in the second half of this year. The company is planning a tabless design to enhance output and will equip the battery with vents at both the top and bottom to allow gas to escape.

LG Energy Solution showcased a UPS battery system using lithium iron phosphate (LFP) batteries, which are known for their excellent fire safety. One section of the company's booth featured a blackout experience zone, allowing visitors to see firsthand how emergency power systems operate during a power outage.

The power driving physical AI, which enables AI to make decisions and take actions autonomously, also comes from batteries. Batteries are installed in humanoid robots and logistics robots. While the market for batteries for robots is still small, it is expected to grow alongside the development of physical AI.

SK On highlighted a logistics robot developed in partnership with Hyundai WIA at its booth, and Samsung SDI displayed both solid-state batteries scheduled for mass production in 2027 and humanoid robots side by side. Above all, what battery companies are most anticipating is autonomous vehicles. Since autonomous vehicles consume large amounts of electricity, they require batteries with much greater capacity.

As the AI era unfolds, battery applications are expanding, but so are expectations. Yongrak Joo, Head of the Technology Research Institute at Samsung SDI, stated, "Batteries for robots must last for more than eight hours and have high output. Because robots are with people 24 hours a day, they require even greater safety than electric vehicles."

K-battery companies are facing challenges due to the electric vehicle chasm, but they cannot stop research and development if they want to secure a competitive edge in the future market.

This content was produced with the assistance of AI translation services.