Skip to Main Content
Advances in sensing technologies and aircraft data acquisition systems have resulted in generating huge aircraft data sets, which can potentially offer significant improvements in aircraft management, affordability, availability, airworthiness and performance (MAAAP). In order to realise these potential benefits, there is a growing need for automatically trending/mining these data and fusing the data into information and decisions that can lead to MAAAP improvements. Smiths has worked closely with the UK Ministry of Defence (MOD) to evolve Flight and Usage Management Software (FUMS™) to address this need. FUMS™ provides a single fusion and decision support platform for helicopters, aeroplanes and engines. FUMS™ tools have operated on existing aircraft data to provide an affordable framework for developing and verifying diagnostic, prognostic and life management approaches. Whilst FUMS™ provides automatic analysis and trend capabilities, it fuses the condition indicators (CIs) generated by aircraft health and usage monitoring systems (HUMS) into decisions that can increase fault detection rates and reduce false alarm rates. This paper reports on a number of decision-making processes including logic, Bayesian belief networks and fuzzy logic. The investigation presented in this paper has indicated that decision-making based on logic and fuzzy logic can offer verifiable techniques. The paper also shows how Smiths has successfully applied fuzzy logic to the Chinook HUMS CIs. Fuzzy logic has also been applied to detect sensor problems causing long-term data corruptions.
Date of Conference: 26-28 June 2005