You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A transformative approach to manufacturing optimization, focusing on the textile forming process. This research synergizes domain-specific knowledge with simulation modeling and introduces Bayesian optimization for efficient parameter space exploration.
Research companion: Heuristic scheduling strategies for single-reactor pharmaceutical batch production under uncertainty. Statistical analysis (JASP) + visualization (Python). Dataset, scripts, and JASP files included.
This project demonstrates an MVP (Minimum Viable Product) for adaptive experimental design that efficiently searches for manufacturing conditions maximizing material fatigue lifetime. Using real CCFatigue dataset, we validate the feasibility of Bayesian optimization for material lifetime prediction.