Our Partners
Co-building scenario-driven research with leading industry players, deploying algorithms onto real production lines
Industry-Research Partners
Our Partners
Every partnership represents real technology deployed on a real production line
A selection of partner companies shown for illustration; all logos remain the property of their respective owners and are used solely to indicate partnership
Case Study · Xiaomi Ecosystem
Die-Casting Quality Foundation Model
Online Die-Casting Quality Monitoring Driven by an Industrial Time Series Foundation Model
Built for online quality perception needs on Xiaomi-ecosystem die-casting production lines, moving from "post-hoc sampling" to "full online inspection." View Project Details →Real Pain Point
Mechanical performance of large die-cast parts relies on destructive testing — every tested part is scrapped, leaving very few labeled samples
Technical Approach
Adversarial semi-supervised training driven by an industrial time series foundation model, combining a small number of destructive-test samples with massive online production-line sensor data
Outcome
Achieved 91% prediction accuracy for comprehensive mechanical performance across regions, with predictions accompanied by reasoning chains and process optimization recommendations
Case Study · General Motors / CITIC Dicastal
Die Fault Diagnosis System
Die Fault Diagnosis System for Automotive Component Manufacturing
Bringing a diagnostic framework for varying operating conditions and long life cycles to the production line, turning online die-state perception into a production-grade software-hardware system. View Project Details →Real Pain Point
The die is the "teeth" of a stamping line — a single fault stops the entire line, and diagnostic accuracy degrades over long-term operation under varying conditions
Technical Approach
A long-life-cycle equipment fault diagnosis framework for varying operating conditions, paired with tailored offline/online sampling strategies
Outcome
Deployed on production lines at General Motors, CITIC Dicastal, and other companies; the research direction has produced 4 high-level SCI publications and 5 granted invention patents
Collaboration Models
How We Partner with Companies
From joint R&D to technology transfer, offering multi-tiered, sustainable industry-academia-research collaboration paths
Joint R&D Projects
Forming joint teams to develop and iterate algorithms around a company's specific real-world production-line challenges
Open Technology Challenges
Companies post key technical needs and the lab tackles them through dedicated projects, such as the Xiaomi Young Scholar Open Challenge Program
Co-Trained Student Internships
Students intern on-site and receive resident technical guidance, refining theoretical methods under real operating conditions
Technical Consulting & Technology Transfer
Providing technical consulting based on existing research, driving productization of algorithms and systems on the production line
See collaboration models, case studies, and process →
Interested in an industry-academia-research partnership? Get in touch
Email: fengjsh7[at]mail[dot]sysu[dot]edu[dot]cn | Address: Room 518, East Building, Science Park, Sun Yat-sen University, 66 Gongchang Road, Guangming District, Shenzhen, Guangdong Province