Dynamic multistep uncertainty prediction in spatial geometry

Maintenance procedures for complex engineering systems are increasingly determined by predictive algorithms based on historic data, experience and knowledge. Such data knowledge is accompanied varying degrees of uncertainty which impact equipment availability, turnaround time unforeseen costs throughout the system life cycle. Once quantified, these uncertainties call robust forecasting to facilitate dependable maintenance costing ensure availability. This paper builds theory spatial geometry as a methodology forecast where available insufficient application traditional statistical analysis. To continuous accuracy, conceptual dynamic multistep prediction model presented applying with long-short term memory (LSTM) neural networks. Based in MATLAB, this deep learning predicts in-service given system. The further into future predicts, lower confidence prediction. Forecasts therefore also made single step ahead. When reached real time, next used update long range here contributed an aggregation quantitative qualitative, subjective expert opinions additional traits such environmental conditions. It beneficial indicate factors prompts greatest aggregated each point. Future work will include option simulate interpolate input enhance accuracy LSTM explore suitable approaches mitigate, tolerate or exploit through learning.