TELD launches AI large model for battery safety, charging

Shanghai (Gasgoo)- On August 15, during the Beijing International Charging Pile and Battery Swapping Station Exhibition 2024 (“CPSE”), TELD, a leading public charging network operator in China, launched a new charging safety and intelligent operations solution based on an AI large model, which the company touted as the country's first battery safety large model built on the charging data from extensive range of vehicles.

As the largest charging network operator in China's public charging sector, TELD operates over 600,000 charging terminals nationwide, providing services to more than 20 million new energy vehicle (NEV) owners. With vast historical data spanning vehicles, batteries, users, terminals, stations, and energy systems, TELD harnesses the computational power and learning capabilities of large models to deeply analyze user charging and discharging behaviors, as well as the true operational characteristics of vehicles. This has enabled TELD to develop a multi-indicator, multi-dimensional battery safety model that offers significant value in safety protection, safety assessment, and health evaluation.

According to data from China’s Ministry of Public Security, by the end of June, 2024, China had gotten 24.72 million NEVs running on roads, accounting for 7.18% of the country’s total automobile population. Of these, 18.13 million units are pure electric vehicles, making up 73.35% of the NEV total. Statistics from the China Electric Vehicle Charging Infrastructure Promotion Alliance shows that as of the same period, China had deployed 10.24 million charging piles, a year-on-year soar of 54%, and 3,772 battery swapping stations.

Mr. Ju Qiang, TELD’s Chief Scientist and Chief Information Security Officer, shared that TELD's network now covers data from 95% of the vehicle models available in the Chinese market, servicing nearly 25 million NEVs across 600 brands. The company has amassed over 100 billion core battery data points. Over its 10-year history, TELD has encountered numerous challenges related to charging safety, station operations, and energy management, underscoring the need for solutions driven by large-scale data and AI models.

Mr. Ju explained that TELD's data encompasses four key areas: batteries, users, charging piles and stations, and vehicles. This includes metrics such as voltage, current, state of charge (SOC), battery management system (BMS) strategies, charging and swapping frequencies, and operational types. He noted, "Data presents a certain pressure for us, with storage costs being particularly high. If data collected over ten years cannot be utilized effectively, it merely becomes a cost burden."

The battery safety large model developed by TELD integrates these four data areas across multiple dimensions and indicators. Currently, the model operates with over 20 billion parameters, capable of identifying 256-dimensional features related to power battery safety protection.