IN 2010, when builders were excavating the site of the former World Trade Center in New York, they stumbled across something rather unusual: a large wooden boat, later dated to the 1700s.
Hitting archaeological remains is a familiar problem for builders, because the land they are excavating has often been in use for hundreds, if not thousands, of years.
Democrata, a UK data analytics start-up, wants to help companies guess what’s in the ground before they start digging. Using predictive algorithms, their new program maps where artefacts might still be found in England and Wales, in order to help companies avoid the time and cost of excavation. “It’s an expensive problem to have once you’ve started digging,” says Geoff Roberts, CEO of Democrata.
Archaeological services can amount to between 1 and 3 per cent of contractors’ total construction cost. “We wanted to bring data science in as an added tool, so humans involved in the process could use it to understand what would likely be found,” says Roberts.
The Democrata team scoured documents from government departments such as the Forestry Commission, English Heritage and Land Registry to find out what the land was used for in the past, for example, and about known archaeological sites. This included “grey literature”, the massive set of unpublished reports written by contractors every year.
With the aid of a supercomputer, they developed models that can pinpoint where treasures are likely to be hidden underground. For instance, land close to water, tin mines or sites of religious significance was ranked more highly than land elsewhere. Other factors like the local geology, animal and plant life also contributed to the score.
This week, Democrata will present the program to engineering companies and the government to hear their feedback. . . . .