Multi-variate time-series for time constraint adherence prediction in complex job shops

One of the most complex and agile production environments is semiconductor manufacturing, especially wafer fabrication, as products require more than several hundred operations remain in Work-In-Progress for months leading to job shops. Additionally, an increasingly competitive market environment, i.e. owing Moore’s law, forces companies focus on operational excellence, resiliency and, hence, leads product quality a decisive factor. Product-specific time constraints comprising two or more, not necessarily consecutive, ensure at level thus, are industry-specific challenge. Time constraint adherence utmost importance, since violations typically lead scrapping entire lots deteriorating yield. Dispatching decisions that determine state art performed manually, which stressful error-prone. Therefore, this article presents data-driven approach combining multi-variate time-series with centralized information predict probability fabrication facilitate dispatching. Real-world data analyzed different statistical machine learning models evaluated.