Batch process data are essentially three-dimensional.Most existing methods on batch process monitoring such as multiway partial least squares(MPLS)and multiway principal component analysis(MPCA)are based on unfolding procedure which represents three-way batch data as a vector in high-dimensional space.As a result of destroying data structures,these methods may lead to information loss.In this article,batch data are considered as a third order tensor and HOPLS is introduced to deal with the data directly instead of performing unfolding procedure.A HOPLS-based online monitoring approach is developed and two new statistics: HO-SPE and HO-T2 are constructed for fault detection and diagnosis.The effectiveness of this approach is illustrated by a benchmark fed-batch penicillin fermentation process.The comparison of monitoring results shows that the proposed approach is superior to MPLS,and it can achieve accurate detection of various types of process faults occurring in the batch operation.