Racket

Rare Class Learning and Unknown Events Detection for Flexible Production
Duration: 01.09.2020 – 31.08.2024

The RACKET project approaches the problem of detecting infrequent and unknown faults through a combination model-based operations and machine learning in theoretical and representative application scenarios. 

The approach is based on the assumption that a physical or process engineering model of a manufacturing plant exists, but it is not fully specified and uncertainties in structure, parameters and variables remain.