An economic analysis of false alarms and no fault found events in air vehicles

Abstract

False Alarms (FAs) that occur in a fielded system and No Fault Found (NFF) events that are discovered after line replaceable units (LRUs) have been returned to repair are costly situations whose full impact is difficult to put into monetary terms. For that reason, pragmatic economic models of NFFs are difficult to develop. In this paper, we deal with the problem of having to differentiate between NFFs of good units under test (UUTs) and of faulty UUTs. While we cannot tell which UUT is good and which is faulty, we can determine using probabilities what percentage of the NFFs are faulty and what percentage are good. Based on these probabilities, we can evaluate various strategies. Assigning cost factors that are knowable, such as the cost of testing a UUT, the cost we incur for good UUTs vs. costs we incur for faulty UUTs and various test and repair costs, we can calculate the performance of various strategies and assumptions. In this paper, we formulate three strategies: 1) We assume all NFF UUTs are good and are willing to endure the cost of bad actors (i.e. faulty UUTs) sent back to the aircraft. 2) We assume all NFF UUTs are faulty and we environmentally stress all NFF UUTs, hoping to fix some and avoid bad actors. 3) We rely on the technician to reasonably select some NFF UUTs and perform appropriate repair. We formulate each of these strategies for a case when NFF is 70%. The formulation is similar with any NFF distribution, but the coefficients in each formula will be different. With proper cost data, we can actually decide which strategy works best. We conclude by tabulating the formulas and calculate NFF costs for an example situation. The numbers we picked for this example may be appropriate for some operations, but not for others. As a follow-up to this paper we would like to validate the model with real data, which may be available in some military and commercial avionics maintenance departments.