27.1 Introduction

Suppose there are an unknown number v of faults in a computer program. A fault is some defect in the code which (under at least one set of conditions) will produce an incorrect output, i.e. a failure. The program is executed for a total time ? under a range of conditions to simulate the operating environment. Failures occur on N _? occasions at times 0< t _{1 2} < < < ?. The time until the detection of fault i is a positive random variable with hazard function ? _i (t) ( i= 1, 2, , v). The first column of Table 27.1 gives a set of failure data for an information system (seconds CPU time). The data can be found in Moek [1]. The main purpose of this study is to estimate the total number of remaining faults in the system.