Workers in manufacturing settings experience highly variable musculoskeletal loading, which current risk assessment methods often fail to fully capture. This study evaluated a Fatigue Failure-Based framework for estimating continuous lumbar loading from variable occupational loads. Worker movements and postures were recorded using Inertial Motion Capture technologies, and L5/S1 joint loading history was estimated through inverse dynamics. Stress cycles were analysed using Rainflow analysis, adjusted with Goodman's method, and summed using Palmgren-Miner rule to estimate cumulative damage. The framework was tested in live industrial settings with eight automotive workers across 108 trials. Logistic regression models demonstrated significant correlations between cumulative damage estimates and self-reported low-back pain (OR = 2.16, 95% CI: 1.30, 3.57). This framework provides a novel method for analysing highly variable loading to estimate cumulative exposure in ergonomics, offering a starting point for future research and potential applications in assessing low back injury risks in similar occupational settings.