Research Fellow (Industrial AI for Manufacturing Process Optimization)

The NTI-NTU Corporate Laboratory seeks a Research Fellow to develop and validate industrial AI methods for adaptive manufacturing process optimization. This role focuses on data-driven approaches combining machine learning, state estimation, and constrained optimization to address process drift and equipment variability. Candidates require a PhD in a relevant field, strong Python and ML framework skills, and experience in areas like time-series modeling, state estimation, or uncertainty-aware modeling.

• Develop machine-learning models for process prediction, state estimation, adaptive optimisation and closed-loop decision support.
• Build structured and reproducible data pipelines for heterogeneous process, sensor, equipment, production and quality data.
• Develop robust methods for process drift, sparse/noisy data and equipment variability, including uncertainty-aware modelling and transfer learning where appropriate.
• Design constrained optimization and sequential decision strategies under engineering limits and human review.
• Validate methods rigorously and translate research outcomes into prototypes, technical reports and publications.

• We regret that only shortlisted candidates will be notified.
• PhD in Computer Science/AI/Data Science or a relevant engineering or materials discipline.
• Strong background in machine learning/data science, with hands-on Python and experience with modern ML frameworks.
• Experience with one or more of time-series/sequence modelling, state estimation, constrained or Bayesian optimization, uncertainty-aware modelling, transfer learning or adaptive experimentation is highly desirable.
• Experience with heterogeneous engineering or manufacturing data and rigorous model validation.
• Manufacturing-process experience is advantageous
• coating/thin-film or run-to-run process knowledge is a plus.
• Strong research record, independent problem-solving ability and effective multidisciplinary communication skills.