| Data scope | Fragmented departmental safety reports, audits, and accident investigations | Enterprise-wide integration of QAR/FD, ACARS, maintenance history, weather, air traffic control, documents, reports, etc. |
| Analysis method | Descriptive statistics, trend analysis, and retrospective root-cause analysis | Anomaly detection, machine learning risk prediction, LLM text analysis, and system dynamics simulation |
| Time orientation | Reactive, partly preventive | Predictive and prescriptive |
| Decision support | Static indicator dashboards and manual reports | Conversational natural-language search, personalized briefings, generative AI reports, and predictive maintenance alerts |
| Organizational culture | Regulatory compliance and mandatory reporting | Just culture-based trust and a virtuous cycle of data contribution and benefit |
| Key limitations | Data silos, lagging indicators, and limited text analysis | Requires integrated technology, organization, and governance, and requires empirical validation |