[Science World] Essential 'Battery Life Management Service' in the Electric Vehicle Era

"Electrification" is now an inevitable trend. Cars, motorcycles, and even ships and airplanes have entered an era where they operate using electric motors. One key issue to address here is battery management. Anyone who has used an old smartphone knows the experience of thinking the battery has enough charge, only to have the call drop unexpectedly. If a car on the road experiences such an issue, it can be fatal to safety.

To solve this problem, research is underway on "battery life prediction" technology. The idea is to equip electric vehicles with devices that can measure how much longer the battery can be used.

The battery's lifespan, or "remaining capacity," is influenced by various conditions such as current, temperature, and usage time. Of course, battery life prediction methods have existed before. These methods organize such conditions into formulas, repeatedly conduct physical tests, and create numerical tables by mathematically modeling the system's degradation process based on those values. During operation, the measured current values are compared to these tables for prediction. This is commonly called the "physics-based prediction method." It works reliably and has minimal errors initially, but it cannot reflect variables arising from different driving patterns. In other words, the prediction error tends to increase over time compared to actual usage.

Therefore, a method often used is to continuously measure the battery current while operating the vehicle and analyze the measurement data to predict the battery's condition. This is a statistical analysis based on data, and nowadays, artificial intelligence (AI) is primarily used for this analysis.

On top of this basic principle, efforts are being made to maximize efficiency by introducing various advanced technologies. A recent notable example is the application of "Digital Twin" technology in battery life prediction research. This method transfers a real electric vehicle into a high-performance computer (HPC) environment to repeatedly conduct experiments in virtual reality, enhancing accuracy. In other words, the shortage of "data," a drawback of AI methods, can be overcome by creating countless driving data through experiments that vary charging and discharging conditions, driving habits, operating conditions, parking and driving environments, and more within virtual reality.

Thanks to these technological developments, battery pilot management services have recently emerged. In the case of a domestic company, a separate terminal is attached to the vehicle, and AI analyzes the battery status and informs the user. Through a smartphone application (app), it provides various figures related to battery usage such as remaining battery value, the ratio of slow to fast charging, energy efficiency, and management ranking, as well as know-how to extend battery life, greatly aiding safe vehicle management.

As such, "Battery Management as a Service (BAAS)," which diagnoses and informs electric vehicle owners about battery status and remaining life, is expected to become an essential service in the electric vehicle era. Since the battery accounts for more than 40% of the electric vehicle's cost and is a core component, it holds great value for vehicle performance and used car price retention. Above all, managing battery performance is a vital requirement for safety. Technologies that guarantee driver comfort, ensure safety, conserve valuable battery resources, and further protect the environment are gaining attention as essential values representing the future automotive culture.

Jeon Seung-min, Science and Technology Writer