Risk analysis in project management

Today's engineering problems are often complex and in real life, however, things do not always go as planned. Duration estimates are just that: estimates. There is always uncertainty associated with events in the future. Also, efforts are not always successful. When something fails, such as a test, some replanning must generally be done, and the results of that replanning usually involve a change to the schedule. With typical schedule analysis, it is difficult (or impossible) to predict the effects of such occurrences. That is why risk analysis is needed. Whereas typical schedule analysis attempts to answer the question "When will the project be complete if all goes as planned?", risk analysis attempts to answer the much broader question "When will the project most likely be finished, given that there are uncertainties to consider?".(It also attempts to answer the same questions about cost.) Risk analysis is performed to focus management attention on the elements of a project that need such an attention in order to permit to attain the wanted goals (technically, costly and in time).

Another way of looking at risk analysis is as a decision support tool, answering the question "What are the possible outcomes of this project?". This is useful to "prove" to his superior or investor how safe it could be to go in this project, when all the risks have been identified, considered and eventually minimized. This last point is in fact why most of the time risk analysis is used. It is performed on three main areas: market-risks, technology-risks and sponsor-risks respectively defined as the acceptability of the concept, the reliability and maintainability of the process, and the integrity and commitment of the developer.

So risk analysis is a logical method for arriving at conclusions about a project and a rigorous approach for dealing with — and quantifying — its uncertainties inherent. It enables to evaluate cases for settlement much earlier, to reduce the cost, to consider alternative resolution, and to prepare budgets and resources.

If a project involves activities that could either succeed or fail, or could follow one of several logical paths, then more thought is required. In these cases, a plan must be developed for each outcome, an index realistically assigned to each one depending of risk gravity and its probability to order the many issues to consider whose results are highly uncertain. Of course the earlier the issues need to be assessed, the greater the uncertainty — and the tougher the task.