Reliability Digital Twin (RDT) is AI application for reliability and maintenance demand.
RDT characterizes how fleet usage leads to failures that drive removals and repairs. Unique features of RDT AI include:
- AI/ML recovery of any reliability distribution, including Weibull and Bathtub (which cannot be modeled by Weibull). The models are understood by Reliability Engineers.
- Rigorous analysis based on part ages, usage, and reliability distribution. Probability distribution of different metrics represented by quantiles, e.g., 5%, 95%, and median.
- Statistical Process Control (SPC) for bad actor detection at given level of statistical significance.
- Maintenance demand trending for line and depot workload, forced (failure) and other removals, No Fault Found load, recurring work on poorly fixed failures, and more.
- Automated monitoring of performance issues associated with identifiable subpopulations of parts. For subpopulations like parts after special repairs, parts modification, or vendor change, the AI reports if reliability improved or got worse.