Modernization Doctrine

Air Force Engineering University Instructor Song Yafei: Where You Put Your Time, There You Will Find Your Harvest

空军工程大学教员宋亚飞：时间用在哪里，收获就在哪里

PLA Daily (解放军报) 26 May 2026

Summary

Song Yafei, an associate professor at the Air Defense and Anti-Missile College of Air Force Engineering University, led a team that from 2017 onward collected tens of thousands of field data sets from frontline units, broke a core data-scarcity bottleneck in 2020, and has since moved an intelligent assisted recognition system into trial use with a PLA unit while sharing the underlying dataset across military teams. The article documents the 'last mile' conversion problem — translating academy research into fielded capability — that PLA military education institutions have repeatedly identified as a structural weakness, and records one institutional workaround: direct researcher-to-unit data-sharing to bypass formal transfer pathways. The framing around 作战辅助决策 (combat-assisted decision-making) and making equipment 'understand adversary intent' points to ongoing work on cognitive-layer AI applications for air defense and anti-missile missions, though the specific system, the units involved, and the current trial-use scope remain unidentified.

Translation

"Where You Put Your Time, There You Will Find Your Harvest"

■ Zhu Qiwei, Zhou Zisha

Air Force Engineering University instructor Song Yafei. Photo by Yang Wenhao

Inside a laboratory at the Air Defense and Anti-Missile College of Air Force Engineering University, at the acceptance testing site for a certain system, Associate Professor Song Yafei was fixed intently on the data flickering across the screen.

"The algorithm recognition accuracy rate has hit a new high! The acceptance test has passed smoothly!" As the cheers of the research group members shattered the silence, Song Yafei, as project lead, was overcome with excitement. This was not only a technical breakthrough — it was yet another concrete result of a military scientific researcher's efforts to "make equipment 'understand' the adversary."

Turn the clock back to 2017, when the team led by Song Yafei was mired in problems related to intelligent recognition. Between "seeing clearly what a target is" and "seeing through what a target intends to do" lay a chasm stretching from perception to cognition.

At a situation analysis meeting heavy with tension, Song Yafei said firmly to his team members: "We who do scientific research must cut roads through mountains and build bridges over rivers. The scarcity of data is precisely proof of its strategic value to the nation and the military! This hard bone — we must gnaw through it!"

In the years that followed, Song Yafei led the team on repeated trips to frontline units across the country to conduct research, successfully collecting tens of thousands of sets of measured data. In the winter of 2020, the team finally broke free from the predicament of data scarcity and solved the core problems of the algorithm model.

Yet Song Yafei did not stop there. He understood clearly that the real battlefield is by no means on paper — only by turning one's gaze toward the exercise ground can the path of scientific research be made steady and far-reaching.

For military academy instructors, the "last mile" of converting scientific research results into real combat capability is both the most critical and the most arduous. With no ready-made conversion pathway available, Song Yafei's team fell into a new predicament of "no shortage of results, but nowhere to take them."

"Waiting, relying on others, and demanding handouts will never bring a breakthrough — you must take the initiative and strike, only then can you break the deadlock." Song Yafei led his members to press forward in the face of difficulty, charging at one "high ground" after another. The lights burning without rest in the laboratory bore witness to their days and nights of struggle for national defense research. Scheme design, model construction, algorithm optimization, simulation verification — every single link had to be refined repeatedly. From data cleaning to model training, from intelligent algorithms to system integration and debugging, Song Yafei consistently pursued the utmost and maintained strict control, letting no critical detail that could affect real-combat effectiveness slip by.

Pressing hard through the assault, they finally broke through the barriers. With the support of the university, the intelligent assisted recognition system developed by Song Yafei's team successfully incubated a key scientific research project and smoothly entered the trial-use phase. In addition, they shared the core dataset with relevant military unit teams, eliminating data barriers and contributing greater combined strength to the cause of strengthening the military.

In recent years, the wave of unmanned systems has swept the globe and the form of warfare (战争形态) has undergone profound transformation. Song Yafei has led his team in sustained breakthroughs in the areas of intelligent command and control, target recognition, and combat-assisted decision-making (作战辅助决策), achieving a number of important results.

Embedding algorithms into national defense construction, using the power of scientific research to safeguard the peace of the homeland. Through nearly ten years of perseverance, Song Yafei has proven a simple truth — "Where you put your time, there you will find your harvest."

Original Chinese

“时间用在哪里，收获就在哪里” ■朱启维周子莎空军工程大学教员宋亚飞。杨文浩摄空军工程大学防空反导学院实验室内，某系统验收测试现场，宋亚飞副教授正紧盯着屏幕上跳动的数据。 “算法识别准确率再创新高！验收测试顺利通过！”当课题组成员们的欢呼声划破寂静，作为项目负责人的宋亚飞内心激动不已。这不仅是技术的突破，更是一名军队科研人员努力“让装备‘读懂’对手”的又一现实成果。时钟回拨到2017年，当时宋亚飞带领的团队，正深陷智能识别相关难题。从“看清目标是什么”到“看透目标想干什么”，中间横亘着从感知到认知的鸿沟。气氛凝重的形势分析会上，宋亚飞对团队成员们坚定道：“我们做科研的，就是要逢山开路，遇水架桥。数据的匮乏，恰恰证明着它对国家对军队的战略价值！这块硬骨头，我们必须啃下来！” 此后几年，宋亚飞牵头带队，辗转祖国各地的一线部队调研，成功收集数万组实测数据。2020年冬，团队终于摆脱数据匮乏的困境，解决了算法模型的核心难题。然而，宋亚飞并未停下脚步。他深知，真正的战场绝非在纸面上，唯有将目光投向演兵场，科研之路才能行稳致远。对军校教员来说，把科研成果转化为实战能力的这“最后一公里”，最为关键也最为艰难。苦于没有现成的转化路径，宋亚飞团队陷入“成果不少、出路难找”的新困境。 “等、靠、要，是换不来突破的，必须主动出击，方能破局。”宋亚飞带领成员们迎难而上，向一个又一个“高地”冲锋。实验室里长明的灯光，见证了他们为国防科研奋斗的日日夜夜。方案设计、模型构建、算法优化、仿真验证……每一个环节都必须反复打磨。从数据清洗到模型训练，从智能算法到系统联调，宋亚飞始终追求极致、严格把控，不放过任何一个影响实战效果的关键细节。奋力攻坚，终破壁垒。在学校支持下，宋亚飞团队研发的智能辅助识别系统成功孵化重点科研项目，顺利进入试用阶段。此外，他们还将核心数据集共享给部队相关团队，消除数据壁垒，为强军事业贡献更大合力。近些年，无人化浪潮席卷全球，战争形态深刻演变。宋亚飞带领团队在智能指挥控制、目标识别、作战辅助决策等方向持续突破，取得多个重要成果。将算法融进国防建设，用科研力量守卫家国安宁。宋亚飞用近10年的坚守，印证了一个朴素的道理——“时间用在哪里，收获就在哪里”。