gong xiaoxu, general manager of xiqing energy

editor's note: since 2022, the Ministry of industry and information technology, the national energy administration and other industry authorities have taken safety early warning technology as the core technology research and equipment direction of new energy storage. Green energy new media recently interviewed the energy storage active safety early warning technology leader Xiqing Energy Technology Co., Ltd.

Xiqing Energy Technology Co., Ltd. (hereinafter referred to as "Xiqing Energy") was born out of Sichuan Energy Internet Research Institute of Tsinghua University. Since its establishment at the end of 2018, Xiqing Energy has established the goal of applying artificial intelligence to the power system.

Gong Xiaoxu is the general manager of Xiqing Energy. He said that by combining the mechanism model of the power vertical with general artificial intelligence, we can dig deeper into the value of data and significantly improve efficiency.

In terms of application, Xiqing Energy focuses on three areas, including the security and operation and maintenance of energy storage, intelligent energy and carbon management in zero carbon parks, and intelligent operation and maintenance of power grid equipment.

At present, the safety and operation and maintenance of energy storage is the main business of Xiqing Energy, accounting for more than 2/3 of the revenue. Specifically, through the self-developed active safety early warning system, Xiqing Energy can early and more accurately warn the faults and risks of energy storage power stations, improve the safety and online rate of power stations and realize predictive maintenance.

Gong Xiaoxu said that Xiqing Energy has the largest power station operation database in the industry. After seven years of data tracking and labeling, and continuous feedback on the improvement of artificial intelligence model, Xiqing Energy has established a leading position in the field of active safety.

at the 13th energy storage international summit and exhibition (ESIE 2025) held in April last year, the early warning accuracy of the three generations of active safety early warning system released by xiqing energy has reached 94%, and the false alarm rate has been reduced to less than 3%. Up to now (March 2026), Xiqing energy security early warning system and algorithm have been applied to more than 33GWh energy storage power station.

To achieve this, Xiqing Energy does not need to install sensors to collect more data, but uses artificial intelligence models to fully tap the value of the existing operating data of the power station. Xiqing Energy's active safety early warning system can warn of different levels of risks from weeks, hours to more than ten minutes in advance.

In February last year, the National Development and Reform Commission and the Energy Administration jointly issued Document No. 136, and new energy and energy storage entered the market in an all-round way. The value of Xiqing Energy's active safety early warning system in the full life cycle return of the power station is also highlighted.

First, based on early warning of faults, predictive maintenance is realized, online rate is improved, and operation and maintenance costs are reduced. The second is to accurately evaluate the state of charge and health to support better market transactions of power stations.

Based on leading active safety early warning technology, Xiqing Energy is expanding its power station operation and maintenance business and will build a complete set of artificial intelligence solutions from active safety early warning to unmanned intelligent operation and maintenance in the future.

"Just like driverless systems can make new energy vehicles more safely adapt to various complex road conditions." Gong Xiaoxu said: "Our goal is to build a'smart driving' system for energy storage power stations, so that energy storage power stations can operate with higher quality and lower cost."

Green Energy New Media: Do you think active safety warning is a trend in the energy storage industry?

Gong Xiaoxu: Yes. Since 2022, the competent authorities have taken active safety early warning technology as one of the core technology research directions of the new energy storage industry. The current lithium battery energy storage can not achieve complete intrinsic safety, passive protection is important, but the best way is to identify hidden dangers early, and timely operation and maintenance to solve security risks.

Green Energy New Media: What is the current situation of the application of active safety early warning system in energy storage power stations?

Gong Xiaoxu: At present, only some energy storage power stations have applied this technology. Only a few regions and individual enterprises require active safety warning systems in regional standards and enterprise standards. For example, Guodian Investment and Shandong Province.

green energy new media: why is this so?

gong xiaoxu: in the past few years, the industry has mainly been doing the construction of energy storage power stations, but has not paid attention to the operation and maintenance. By now, a considerable number of power stations have been in operation for more than a few months, and many problems have arisen. People began to pay attention to whether there are better means or technology to improve the quality of operation. Generally speaking, people's attention and demand for active safety warning are increasing.

Green Energy New Media: What is the difference between Xiqing Energy's active safety early warning system and other countries?

Gong Xiaoxu: Our technical route is similar to the "visual route" of new energy vehicle automatic driving technology. There is no need to install new sensors. It is mainly based on the collected data types. Do in-depth mining.

xiqing energy active safety early warning system comprehensive research interface

green energy new media: what are the advantages of this technical route compared with other countries?

gong xiaoxu: using the technical route of big data and artificial intelligence technology combined with mechanism model, it can be identified in the hidden danger stage, and minor faults can be alarmed. compared with other technical routes focusing on the early warning of battery thermal runaway, the observation stage is longer and the hidden danger is found earlier.

green energy new media: how do the effects of different technical routes compare?

gong xiaoxu: our active safety early warning system can achieve three levels of protection, from several weeks in advance, dozens of hours in advance to 15 minutes in advance. Especially several weeks in advance, when the battery body is not damaged and a single parameter is normal, the artificial intelligence model can capture nonlinear anomalies, such as inconsistencies become larger, and identify system safety hazards.

green energy new media: but xiqing has also released a self-developed sensor?

gong xiaoxu: the sensor we developed is a supplement to our thermal runaway warning. Thermal runaway warning is done purely with data, and there will be a certain lag. Many batteries will release combustible gases hydrogen and carbon monoxide in the early stage of thermal runaway. We want to enhance our early warning ability of thermal runaway by monitoring these special gases.

At present, the service life of similar sensors in the industry is generally only two or three years. Our sensor adopts the ultrasonic technology route, the life span is increased to more than ten years, and has an earlier warning time, which can be regarded as a pain point for the industry.

Green Energy New Media: In addition to improving safety, is Xiqing's active safety early warning system helpful to improve the return on investment?

Gong Xiaoxu: Yes. In terms of improving safety, once a safety accident occurs, the investment is directly cleared, which is the biggest return. Active safety warning can be linked to predictive active operation and maintenance to avoid battery failure and maintenance. Due to the short board effect of the battery, a faulty cell will drive the deterioration of the state of other cells, timely maintenance and replacement of the battery, can improve the life of the battery life cycle, reduce operation and maintenance costs.

In terms of evaluating the state of the energy storage power station, our system can also accurately evaluate the state of charge and health of the battery. This is also a difficult point in the industry, but higher accuracy can be achieved through artificial intelligence. In this way, the power station side can clearly know the capacity of its own power station and avoid losing value or being punished by the power grid when declaring the output curve because of inaccurate capacity.

xiqing energy active safety early warning system won the first prize of China electric power innovation award

green energy new media: what is the layout of xiqing energy in operation and maintenance?

Gong Xiaoxu: Our operation and maintenance is based on the advantages of predictive operation and maintenance brought by our active safety early warning system. In the long term, our goal is unmanned intelligent operation and maintenance, combining our artificial intelligence model with robots. However, at this stage, we are also considering cooperating with some operation and maintenance companies.

green energy new media: will xiqing energy expand its artificial intelligence capability to other fields?

gong xiaoxu: our barriers to active safety warning are very important because we have been cleaning and marking power station data for 7 years. In the future, it is also necessary to track the operating data of the power station under different operating conditions and different scenarios to further improve the reliability and versatility of our model. At this stage, our goal is to specialize in the field of energy storage active security and to be a small and beautiful company with deep vertical.