Abstract: Battery node diagnostic data may be received from a battery system. One or more outlier detection machine learning models may be selected based on profile information included in the battery node diagnostic data. The profile information may identify a battery node operation profile associated with some or all of the battery node diagnostic data. One or more of the battery nodes may be identified as outliers by applying the one or more outlier detection machine learning models to identify one or more differences between first diagnostic data for the designated subset of the battery nodes and a population-level representation of the battery node diagnostic data. Outcome values may be determined by applying one or more predetermined rules to the battery node diagnostic data. A battery node may be identified as exhibiting a fault based on the designated subset of the battery nodes and the plurality of outcome values.

Abstract: Signal data may be received from battery node controllers corresponding with battery nodes in a battery system operating in a charging or discharging state. A battery node controller may selectively couple a power bus in the battery system with a battery node including a respective one or more battery cells. Candidate events may be identified as anomalous based on a discontinuous changes in the signal data. The candidate events may correspond to a respective battery node and a respective period of time. When one or more candidate events are identified as anomalous, diagnostic data may be sent to a remote computing system over the internet via a communication interface. An indication of a designated battery node as defective based on one or more machine learning models applied to the diagnostic data at the remote computing system may be received. A battery node operation profile associated with the designated battery node may be updated.

Abstract: Battery system diagnostic data may be generated at battery node controllers in a battery system. A battery node controller may selectively couple a power bus with a battery node including one or more battery cells. An unsupervised machine learning model may be trained based on the battery system diagnostic data. Outlier values may be determined by applying the machine learning model to the battery node diagnostic data. A battery node may be identified as anomalous based on a comparison of a designated outlier value associated with the designated battery node with the plurality of outlier values. A battery node operation profile controlling operation of the designated battery node may be updated.