Fast mission reliability prediction for Unmanned Aerial Vehicles

There is currently a significant interest in the use of autonomous vehicles in many industrial sectors. One such example is the ever increasing use of Unmanned Aerial Vehicles (UAVs), particularly in military operations. This enables dangerous missions to be accomplished without risk to a pilot. UAVs also have potential civil applications which would require their certification and the demonstration that they are able to respond safety to any potential circumstances. The aircraft would therefore need to be capable of responding safely to the occurrence of component failures, the emergence of threats such as other aircraft in the neighboring airspace, and changing weather conditions. The likelihood that an aircraft will successfully complete any mission can be predicted using phased mission analysis techniques. The predicted mission unreliability can be updated in response to changing circumstances. In the event that the likelihood of mission failure becomes too high then changes have to be made to the mission plan. If these calculations could be carried out fast enough then the quantification procedure could be used to establish an acceptable response to any new conditions. With a view to using the methodology in the context described above, this paper investigates ways in which phased mission analysis can be improved to reduce the calculation time. The methodology improves the processing capability for a UAV phased mission analysis by taking into account the specific characteristics of the fault tree structures which provide the causes of phase failure for a UAV mission. It also carries out as much of the quantification as possible in advance of the mission plan being formulated.