Key research achievements, competition honors, and industrial application highlights of the laboratory
Major Breakthroughs
Flagship results across our three research thrusts: intelligent process monitoring, online fault diagnosis, and real-time production/maintenance optimization
An operating-condition-aligned, risk-aware RUL prediction framework won first place worldwide — the first champion team from a Chinese university in the history of the competition
A physics-informed hypergraph multi-task network achieves 99.55% compound-fault diagnosis accuracy with selective cascaded inference below 50% routing rate (RESS 2026)
Fusing physical mechanisms with time-series AI: anomalies as small as 0.02mm detected online, batch anomalies reduced by 97.35%, customer-confirmed defect rate down 78%
Industrial Applications
Working with General Motors, Xiaomi, CITIC Dicastal, Applied Materials, SAIC KB, Shanghai Electric, Taiyuan Heavy Industry and more
The Profile Abstract streaming-data method has been deployed at Applied Materials and SAIC KB; online monitoring systems run on real stamping and injection-molding lines
Long-life-cycle diagnosis frameworks under varying operating conditions applied at General Motors, Taiyuan Heavy Industry, and CITIC Dicastal — covering die/mold diagnosis, engine abnormal-noise detection, and servo motor monitoring
Offshore wind farm maintenance scheduling and injection-molding risk assessment systems productized with Shanghai Electric and REthink Technology; cases featured in books such as Industrial Big Data and Industrial AI
Academic Achievements
Research published in authoritative journals of reliability, industrial informatics, and advanced manufacturing (see Google Scholar for the full record)
Uncertainty-informed cascaded diagnosis of compound faults via a physics-informed hypergraph framework (2026)
Profile Abstract: optimization-based subset selection and summarization for profile data mining (2022)
Online virtual metrology with sample selection for slow-drift manufacturing processes (2019)
Active in the PHM community and contributing to IEEE Computer Society smart-manufacturing standards activities
IP & Honors
30+ peer-reviewed publications and 100+ patents granted in China, the US, Germany and beyond
Chinese, US, and German patents in industrial time-series analysis, fault diagnosis, and maintenance optimization
CITIC Group inaugural Science & Technology Special Prize; Industrial Application Innovation Award of the National Engineering Center for Deep Learning; Guangdong Industrial Software Award; PHM Society Doctoral Scholarship
PHM international data challenge world champion; PI recognized as Shenzhen High-level Overseas Talent (Peacock Plan)
The SAM-IPA-1 team from AI Cube Lab — led by Prof. Jianshe Feng and composed of undergraduate, master's, and doctoral students Peng Gao, Fanyu Qi, Yizhang Zhu, Jianyu Zhang, and Wenfei Li — leveraged outstanding technical expertise and an innovative algorithmic architecture to outshine international competitors and claim first place at the PHM 2025 Data Challenge.
The AI Cube Laboratory team (SAM-IPA-1) defeated numerous strong teams from around the world to win the PHM North America Data Competition championship, becoming the first Chinese university team to win this competition. Notably, the team's main force is first-year master's student Gao Peng, who just enrolled.
Prof. Jianshe Feng has been selected for the "Emerging Leaders 2026" special issue of the internationally authoritative journal Measurement Science and Technology, in recognition of his outstanding contributions to multimodal spatiotemporal data modeling and industrial AI, demonstrating his growing academic influence as an emerging leader in the field.
Welcome to visit our research projects page, or contact us directly for in-depth communication
