Unified accelerated life testing for one-shot devices with Weibull lifetime distributions

In traditional testing paradigms, step-stress accelerated life testing (SSALT) employs a single test group subjected to a single stress-change pattern, while constant-stress accelerated life testing (CSALT) utilizes multiple test groups exposed to uniform stress levels without variation. This article introduces a novel testing methodology, termed unified accelerated life testing (UALT), which integrates the principles of CSALT and SSALT. The UALT framework incorporates multiple test groups, each subjected to distinct stress-change patterns. By applying multiple stress-change patterns to test items, the UALT framework enhances data collection efficiency and provides a more accurate reliability estimation. The study conducts a detailed comparative analysis of these three optimal designs. Simulation results demonstrate that UALT with step-down-stress patterns achieves superior efficiency in data collection, as evidenced by smaller bias and standard deviation of reliability estimates. This innovative approach offers a nuanced understanding of how diverse stress-change patterns influence product reliability assessment. To illustrate its practical applicability, this article presents a real-world case study involving grease-based magnetorheological fluids, showcasing the effectiveness of optimal UALT design in reliability testing. Copyright © 2025 IEEE.