Autonomous fault detection and diagnostics, an enabler to control risks of military operations

This paper aims to reveal how the Netherlands Armed Forces may mature in autonomous fault detection and diagnostics. Generally, fault detection and diagnostics is important as lacking asset health may amplify the risks of military operations. Autonomous fault detection and diagnostics could make an individual’s knowledge explicit and it could be essential to process the expanding amount of data from (health) monitoring systems. The state-of-the-art in autonomous fault detection and diagnostics will be projected on a realistic case study. It will be shown that the autonomous fault detection and diagnostics in this typical case study benefits from model-based limits that were already implicitly used. Model-based limits are known to better cope with the varying and unprecedented nature of military operations than value-based limits. The underlying problem is that the risks assigned to faults tend to evolve more rapidly than the designed-in autonomous fault detection and diagnostics. The challenge to make autonomous fault detection and diagnostics robust against varying risk assessments remains unresolved.