[4L2-GS-13] AI application: Anomaly detection and maintenance

[4L2-GS-13-04] Deep Learning Image Recognition for Jet Engine Maintenance

〇Hitoshi Hattori¹, Kazuo Yonekura¹, Masaya Kuribara¹, Masanori Hasegawa¹, Nozomu Hashizume¹, Koji Tokunaga¹ (1. IHI corporation)

Keywords:Jet Engine , Maintenance, Image Recognition

We have been propelling AI researches to create new value for jet engine maintenance service. Our first attempt is applying AI to engine parts damage recognition to reduce workload. Since engine parts damage information is necessary for customer reports, a large number of images have to be sorted in a correct order. However, it is sorted manually and it takes half a day for one report. In this paper, we present image classification system for the engine parts.