International Journal of Science, Engineering and TechnologyAuthors: Mr. Harunmiya Sirajmiya Malek
Abstract: One of the most difficult problems affecting the dependability and security of contemporary Windows operating systems is still kernel-level failures. The Critical Process Died error is the most important of these errors since it causes a Blue Screen of Death (BSOD) when a crucial operating system process fails, abruptly ending system execution. In addition to disrupting regular computer operations, these failures also affect cloud platforms, enterprise services, AI workloads, and digital security infrastructures. Recent developments in AI-driven system monitoring have shown a great deal of promise for spotting unusual system behavior before disastrous catastrophes take place. Recent studies on log-based anomaly detection, graph neural networks, transformer-based language models, contrastive learning techniques, AIOps frameworks, and eBPF-enabled observability for predictive system monitoring (2022–2024) are all critically examined in this study. The study also evaluates current methods according to their capacity for detection, computational effectiveness, scalability, and suitability for Windows-based settings. In order to determine future paths for intelligent operating system reliability, current research trends, constraints, and unresolved issues are examined. Lastly, a hybrid AI-driven framework that incorporates anomaly detection, kernel telemetry, and log analytics is suggested to strengthen the resilience of next-generation computing systems and improve early failure prediction.
DOI: http://doi.org/10.5281/zenodo.21426385
