KAIST develops ultra-precise battery inspection technology to prevent EV fires

South Korean researchers have developed an ultra-precise non-destructive inspection technology capable of detecting defects in electric vehicle batteries during manufacturing — without disassembling or damaging them.

KAIST announced Thursday that a research team led by Kim Young-jin, a professor in the Department of Mechanical Engineering, has developed a technology that can measure the thickness of lithium-ion battery electrodes in a non-contact, non-destructive manner, detecting differences as small as one ten-thousandth the width of a human hair.

The electrodes in lithium-ion batteries, widely used in electric vehicles, are critical components through which electricity flows. Even a slight variation in electrode thickness can cause current to concentrate in specific areas during charging and discharging, generating heat. If that heat accumulates, it can trigger thermal runaway — making uniform electrode thickness management essential in battery manufacturing.

Existing inspection technologies, however, have struggled to keep pace with production lines. X-ray computed tomography can examine internal structures in detail, but the lengthy scanning time makes it impractical for fast-moving manufacturing environments.

The research team addressed this problem by combining an optical frequency comb with THz waves. The technique works by directing THz waves at a battery electrode and capturing the signals produced as the waves reflect multiple times inside it. Those signals are then analyzed with extreme precision using the optical frequency comb as a reference, allowing the electrode's thickness to be calculated. The approach enables nanometer-level measurement of the electrode's interior without causing any damage.

Central to the technology is its use of Fabry-Pérot interference — a pattern of evenly spaced interference fringes created as THz waves travel back and forth multiple times between the front and back surfaces of the electrode.

As a result, the team succeeded in simultaneously measuring both the electrode thickness and its complex refractive index in a single pass, with no additional calibration required.

The team validated the technology on battery electrodes measuring 50 to 150 micrometers thick — roughly the width of a single human hair. A measurement time of just 0.2 seconds was sufficient to detect thickness variations as small as one fourteen-hundredth the width of a human hair, a level of precision compatible with fast-moving battery production lines.

The research is significant for introducing an inspection method capable of identifying invisible microscopic defects during production without disassembling or damaging the battery. The technology is expected to serve as a key tool not only in lithium-ion battery manufacturing but also in the production of next-generation all-solid-state batteries — which use a solid electrolyte instead of a liquid one — enabling early defect detection and improving both safety and production stability.

"Applying this technology will require demonstrating its robustness in actual production environments," Kim said. "Once it is integrated with an automated control system that feeds measurement results directly into process adjustments, and once the equipment is miniaturized, field deployment will become feasible."

The findings were published in the international journal Nature Communications.