It is trite to observe that subsurface conditions are notoriously difficult to evaluate and predict. Practical limitations on geological and geotechnical investigations mean that actual subsurface conditions will not be fully known until after the construction is underway and in some cases only when the project is completed.

Take as an example traditional geotechnical investigations for tunnelling works. There is little information about actual conditions across a project site that can be gleaned from a finite set of bore logs, cores and test pits typically funded by the principal before a contract is tendered and awarded. Because of the relatively significant distance between locations of bores and where cores have been collected and the random distribution of rocks and other subsurface conditions, the data is limited. The only accurate information is at the location where a bore log is available (assuming that the log accurately records what was found or the cores that can be inspected).