Software Reliability Growth Model

Every test campaign finds bugs fastest at first, then slower and slower as the obvious defects run out — a pattern software reliability engineers model as a non-homogeneous Poisson process. This simulator seeds a testing run by drawing a random total defect count from a Poisson distribution and giving each defect an independent, exponentially-distributed discovery time, then plays the test clock forward: a growing 3D bar chart shows the actual, stochastic cumulative defect count against the smooth Goel-Okumoto mean-value curve it fluctuates around. Adjust the ultimate defect count and detection rate to see how codebase size and test-suite maturity reshape the discovery curve, and watch the live discovery-rate and remaining-defect readouts track a real quality-assurance dashboard.