A*STAR

Research Scientist (Predicative Quality), IPV, ARTC

A*STAR
ResearchSingaporeFull-time2 weeks ago

About the role

AI summarised

Research Scientist role at ARTC focusing on predictive quality for advanced manufacturing. The position involves leading R&D of novel AI architectures fusing vision and temporal data, developing root cause analysis methods, and implementing knowledge graphs with LLMs/VLMs for automated quality sentencing.

ResearchFull-time

Key Responsibilities

  • Lead the research and development of novel AI architectures that fuse vision data with temporal manufacturing process data to predict final product quality.
  • Develop advanced methodologies for Root Cause Analysis (RCA), moving beyond correlation to establish causal links between process variables and inspection outcomes.
  • Design and implement Knowledge Graphs and semantic reasoning layers that integrate domain expertise with LLMs/VLMs to automate 'final sentencing' and provide explainable AI (XAI) insights.
  • Architect and fine-tune state-of-the-art multimodal models to enable text-promptable vision inspection and contextual decision-making.
  • Pioneer the use of Temporal Transformers or Physics-Informed Neural Networks (PINNs) to analyze complex manufacturing time-series data for anomaly detection and yield prediction.
  • Document research in high-impact internal reports or patent filings and stay at the forefront of AI/ML literature to maintain the institute competitive edge.
  • Provide technical oversight for QC/QA governance frameworks and mentor junior engineers in data integrity and model validation.

Requirements

  • Ph.D. in Computer Science, Machine Learning, Electrical Engineering, or a related quantitative field is mandatory.
  • Demonstrated experience in publishing or developing innovative algorithms in Computer Vision, Predictive Analytics, or Multimodal AI.
  • Deep understanding of AI-based image segmentation, classification and time-series analysis and signal processing.
  • Hands-on experience with Knowledge Graphs, ontologies, or graph neural networks (GNNs).
  • Strong background in Root Cause Analysis (RCA) and statistical process control.
  • Advanced Python programming skills.
  • Ability to drive research projects from conceptualization to a deployable 'target product.'
  • Exceptional ability to communicate complex scientific concepts to both technical peers and non-AI manufacturing stakeholders.