특강

Developing Intelligent Systems for
Industrial Scenarios: Experiences, Challenges,
and Future Directions

Sheng Miao
Ph.D. Info. Tech.
Associate Professor
Qingdao University of Technology, China


Abstract:
The transition from laboratory research to deployed intelligent systems in industrial environments presents a distinct set of challenges rarely addressed in academic literature. This talk draws on extensive hands-on experience in designing, implementing, and maintaining AI-driven solutions within the industrial sector, where data, people, and processes intersect under constant operational pressure. The discussion centers on practical hurdles encountered throughout the full project lifecycle. Key topics include the often-underestimated difficulties of data acquisition in legacy facilities—specifically sensor reliability, data quality issues, and the integration of heterogeneous data streams—as well as the critical necessity of process analysis and domain understanding. From a management perspective, the talk shares lessons learned in milestone planning, team composition, and managing stakeholder requirement changes alongside risk mitigation strategies. On the technical front, it addresses principles for model selection under resource and latency constraints, as well as post-deployment operations and maintenance in continuous production environments. Ultimately, it is argued that successful industrial AI requires a holistic approach bridging engineering, domain knowledge, and project management. The session concludes by outlining promising directions for the next generation of robust, adaptive, and maintainable intelligent systems for industry

Biography:
Dr. Sheng Miao is an Associate Professor at Qingdao University of Technology, China, a position he has held since 2021. He received his Ph.D. in Information Technology from Towson University, United States, in 2017. His academic and research trajectory includes roles as an Assistant Researcher at the U.S. Army Research Laboratory (2014), an Assistant Professor at Qingdao University, China (2017), and a Visiting Researcher at the University of Kitakyushu, Japan (2022). Dr. Miao is an active contributor to the international research community, with over 60 peer-reviewed publications to his name and a consistent record of chairing and organizing conferences across multiple venues.
Dr. Miao's research interests center on multimodal data-driven intelligent systems for industrial applications, digital twin technology, deep learning in smart city contexts, ecological informatics, and AI for healthcare. He is particularly known for his hands-on expertise in the full lifecycle of smart systems—from design and implementation to deployment in complex, real-world environments. Over the course of his career, he has spearheaded several high-impact applied projects, including intelligent upgrades for wastewater treatment facilities, an assistive living system for individuals with Spinal Muscular Atrophy, and remote sensing solutions based on satellite and UAV imagery for water quality forecasting and vegetation cover assessment. These systems have been successfully adopted by governments, large corporations, and medical institutions in both China and the United States, with consistently positive feedback from end users.