Data analytics for engineering and construction project risk management

by Ivan Damnjanovic, Kenneth Reinschmidt

Data analytics for engineering and construction project risk management is about time and chance as they affect projects.

More specifically, the main objective of Data analytics for engineering and construction project risk management is to provide foundations for the assessment of uncertainty and risks on engineering projects of all types.

It deals with the spectrum of uncertainty, from the variability in construction operations to the risks in unique, complex, first-of-a-kind projects. In looking at field operations or other project activities, we use probabilities to try to describe the natural variability of work, and we are concerned with answering the question:

Is the reported performance on this activity merely reflecting these inherent variations, or is it sending a message that the activity is about to go seriously out of control?

In looking at major risks on complex projects, we use data, information, and knowledge about the underlying behavior to express the confidence in our risk estimates.

In general, Data analytics for engineering and construction project risk management places emphasis on building data-driven models, and these models are of necessity mathematically inspired.

As the British physicist William Thomson (Lord Kelvin) said (1883), “when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind.” And “meager and unsatisfactory” is a good description of many project risk assessments.

It may be argued that project managers should place their confidence in experience, judgment, and gut feel, not mathematics. However, learning about risk by the trial-and-error method can take a long time, and the lessons can prove to be very expensive.