Projects, pathogens and incubation periods

Errors can have wide-scale effects in large, complex project organisations. Such errors can typically be traced back to latent conditions, or `pathogens', that tend to reside in a system until they become actual failures – often as a result of some distinct incubation process. A study was undertaken to identify and categorise the pathogens and incubation processes that lay behind a set of significant errors occurring in large-scale engineering projects. The most common kind of pathogen arose from practices that organisations or individuals had adopted (for example reusing existing design solutions). The next most common kind arose from inherent properties of the task being performed (for example disproportionate and non-linear relationships). As far as incubation processes were concerned, the most common kind was the simple accumulation of probability: because latent conditions were not detected from one project to the next, the cumulative probability of a failure occurring on some project simply increased over time until an error was virtually inevitable. Other kinds of incubation process included organisations becoming increasingly habituated to circumstances that eventually changed, and the natural accumulation of complexity in various systems. These results should help provide a systematic way of assessing how vulnerable a project organisation is to making significant errors.