Impedance Growth as the Early Warning a Buyer Can Measure
Early impedance growth exposes internal battery interphase degradation long before standard capacity tests reveal physical performance loss.
Z-Score Outlier Detection
Meaning
Statistical screening techniques that measure how many standard deviations a cell’s performance parameters deviate from the production batch mean identify potential defectives. Sourcing quality teams apply this z-score outlier detection to isolate and reject atypical cells before they are assembled into battery packs. This analysis is calculated by dividing the difference between a cell’s parameter value and the batch mean by the standard deviation.
It is not applied to prototype lines or small development runs where the sample size is too small to calculate a statistically valid standard deviation.
Statistical Analysis
High volume cell production produces a natural distribution of performance values for capacity, weight and internal resistance. This z-score outlier detection uses these distribution curves to identify cells that lie far from the average, even if their values fall within the absolute datasheet limits. A cell with a high deviation score, typically greater than three, is flagged as an outlier because its unusual behavior suggests a hidden manufacturing defect.
Sourcing teams use this statistical filter to ensure that only the most consistent, uniform cells are selected for pack assembly.
Manufacturing Screen
Anomalies in the manufacturing process, such as slurry mixing errors, electrode coating variations or tab weld defects, can create cells with atypical properties. This z-score outlier detection runs on the end of line test data to catch these outliers before the cells are packaged for shipment. The screening looks at multiple parameters simultaneously, such as open circuit voltage, self discharge rate and high current resistance.
Catching these defective cells early reduces the risk of cell imbalance and early pack failure during operation.
Yield Optimisation
Procurement contracts specify the maximum allowed percentage of outlier cells in each production lot to enforce high manufacturing standards. This z-score outlier detection is integrated into the supplier’s quality control software to automate the screening process and improve yield tracking. Sourcing agreements include clauses that allow the buyer to reject any shipment that contains an excessive number of outlier cells, even if those cells pass individual tests.
This strict requirement protects the buyer from using unstable cells, securing the long term reliability of the battery system.