🏆 Congratulations: AI Cube Laboratory Wins PHM North America Data Competition Championship, Becoming First Chinese University Team to Win

On September 19th, exciting news came from the Advanced Manufacturing Intelligence Laboratory! The AI Cube Laboratory team (SAM-IPA-1) won the championship in the PHM North America 2025 Data Competition (PHM North America 2025 Conference Data Challenge), standing out among numerous top-tier competing teams from around the world and becoming the first Chinese mainland university team to win this top-tier international competition in history, achieving a historic breakthrough!

The PHM (Prognostics and Health Management) North America Data Competition is one of the most authoritative and influential international competitions in the field of industrial intelligence and predictive maintenance, organized by the International PHM Society. This competition attracted numerous competing teams from top-tier universities, research institutions, and well-known enterprises from around the world, with exceptionally fierce competition. The competition focused on the extremely challenging industrial application scenario of aircraft engine health management (Engine Health Management), requiring participants to accurately predict the remaining cycles of three key maintenance events based on massive sensor data from 8 commercial jet engines (each engine with up to 15,000 data points covering 2,001 flights in different flight phases): High-Pressure Turbine Shop Visit, High-Pressure Compressor Shop Visit, and High-Pressure Compressor Water-Wash. Participants needed to extract effective features from as many as 16 sensor variables and achieve high-precision prediction under a time-weighted error scoring system. This placed extremely high demands on teams’ knowledge of aircraft engines, time-series data analysis capabilities, deep learning modeling skills, and engineering practice abilities.

The SAM-IPA-1 team dispatched by the AI Cube Laboratory performed excellently in the competition with outstanding algorithm design and deep accumulation of industrial artificial intelligence technology. The team deeply analyzed multimodal sensor data characteristics of aircraft engines in different flight phases such as takeoff, climb, and cruise, innovatively proposed intelligent prediction methods integrating deep learning with physics-driven approaches, fully utilizing component degradation patterns caused by engine cycle stress to construct precise remaining useful life prediction models. This method reached internationally leading levels in key indicators such as prediction accuracy, robustness, and generalization capability, successfully addressing the high-penalty challenge of delayed prediction in the time-weighted scoring system, ultimately winning the championship with excellent results.

Particularly noteworthy is that the main force of this championship team is first-year master’s student Gao Peng, who just enrolled. As a 2025 freshman, Gao Peng participated in such a high-level international competition only weeks after enrollment and played a key role. Facing complex multi-phase flight data from aircraft engines, he quickly mastered sensor data characteristics under different operating conditions from ground idle, taxi takeoff, climb, cruise to landing, and deeply understood the physical mechanisms of component degradation caused by engine cycle stress. He demonstrated solid mathematical foundations, keen problem insights, and outstanding programming implementation capabilities in the competition, undertaking important work in core algorithm design and model optimization for the team, especially proposing innovative solutions in handling the time-weighted scoring system and balancing prediction accuracy with timeliness. Gao Peng’s excellent performance fully demonstrates the unique advantages of the AI Cube Laboratory in rapid graduate cultivation—through the cultivation model of “promoting learning and research through competition,” freshmen can quickly enter research states and grow rapidly in practice.

This championship victory is of extraordinary significance. As the first Chinese mainland university team to win the PHM North America Data Competition championship, the AI Cube Laboratory not only brought honor to Chinese universities but also fully demonstrated the strong research capabilities and technical innovation abilities of domestic universities in the field of industrial artificial intelligence. This achievement broke the long-standing monopoly of this top-tier international competition championship by European and American universities and well-known enterprises, marking that Chinese universities have reached internationally leading levels in the field of industrial intelligence and predictive maintenance.

This award reflects important achievements of the laboratory’s long-term deep cultivation in the field of industrial artificial intelligence. Laboratory head Associate Professor Feng Jianshe has deep theoretical accumulation and rich engineering practical experience in directions such as industrial intelligence, predictive maintenance, and multimodal time-series analysis, having engaged in industrial intelligence R&D work for many years at well-known enterprises such as General Motors USA and Foxconn Industrial Internet. In such an academic atmosphere and under supervisor guidance, laboratory students can access the most cutting-edge research topics and most real industrial scenarios, laying a solid foundation for achieving excellent results in top-tier international competitions. It is worth mentioning that this competition started on July 1st, underwent fierce competition for over two months, closed submissions on September 5th, and announced finalist teams on September 9th, with the SAM-IPA-1 team successfully entering the finals and ultimately winning the championship through excellent performance. The team will be invited to present research results at the PHM 2025 International Conference held from October 27-30.

For students aspiring to pursue studies in the field of industrial artificial intelligence, this achievement has important reference value. It fully demonstrates that in the AI Cube Laboratory, even newly enrolled graduate students can quickly master cutting-edge technologies under supervisor guidance, participate in top-tier international competitions, and achieve excellent results. Through systematic cultivation programs, rich project resources, and an open academic atmosphere, the laboratory provides full growth space and demonstration platforms for each member.

This championship victory not only enhanced the AI Cube Laboratory’s visibility and influence in international academia and industry but also demonstrated the laboratory’s outstanding capabilities in intelligent operations and maintenance technologies for critical fields such as aerospace and high-end equipment. Aircraft engine health management is a typical intelligent operations and maintenance application scenario for high-end equipment, involving complex physical processes, massive multimodal time-series data, and strict safety requirements, representing the “crown jewel” of industrial artificial intelligence technology. The laboratory’s technical breakthrough in this field has accumulated valuable experience for future applications of intelligent predictive maintenance technologies in broader industrial scenarios such as new energy vehicles, semiconductor manufacturing, and advanced manufacturing equipment.

This achievement creates favorable conditions for the laboratory to conduct higher-level international academic exchanges and undertake more important research projects. The laboratory will continue to adhere to the development philosophy of “balancing research and engineering, integrating theory and application,” cultivate more high-level talents with international competitiveness, and contribute strength to promoting the development of China’s industrial intelligence technology and intelligent upgrading of high-end equipment!