Reverse-time simulation for predictive defect prevention in automotive stamping industry: A backward planning framework

Automotive stamping is a complex manufacturing process where even minor defects can lead to costly rework, scrap, and production delays. This paper introduces a backward simulation planning methodology to prevent defects by virtually “reversing time” – tracing quality issues back to their root causes and proactively adjusting processes before defects occur. Unlike traditional forward simulations that predict outcomes from given inputs, backward simulation uses desired outcomes (zero-defect targets) as a starting point and works in reverse to identify required process conditions. We integrate this approach into a digital twin of an automotive stamping line, using discrete-event production simulation and high-fidelity forming simulations. A realistic case study of a multi-stage automotive press line is presented, including complete technical calculations and a closed-loop corrective feedback system. We detail system architecture, timeline-reversal algorithms, and the coupling of simulation tools (e.g., AnyLogic for production flow, Siemens Tecnomatix for process simulation) with shop-floor data. The backward simulation identified root causes of defects (such as wrinkles and cracks) by tracing final part defects to specific press settings and material conditions upstream. A comprehensive defect prevention loop was implemented, enabling real-time feedback to adjust press parameters and scheduling. Results show a significant reduction in defect rate (from 4.0% to 1.5%), improved throughput, and cost savings of over 60% in scrap reduction, confirmed by performance metrics. Figure 10 summarizes key improvements This paper’s depth and technical sophistication align with industry white paper standards, offering a structured framework – from Abstract through Conclusion – for deploying backward simulation in JIT automotive stamping. The approach demonstrates how reversing the simulation timeline, combined with real-time data, can effectively eliminate root causes of defects before they manifest, thus ensuring first-time quality in high-volume manufacturing.