[CES 2021] ArchAI: Using AI to automatically detect archaeological sites
CrossMind.ai logo

[CES 2021] ArchAI: Using AI to automatically detect archaeological sites

Jan 12, 2021
Cultures across the world value their heritage and archaeology. Because of this, in the UK, like many countries, it is a legal requirement to assess the potential for archaeological damage when submitting planning applications for building work. Most of us think of Stonehenge when hearing the word archaeology but most archaeology is hidden underground and it’s really challenging to understand where lies what. As a result, currently it takes 6-12 months to make an accurate archaeological assessment. Throughout this period more planning costs are sunk into a planning application. During this assessment they might come across archaeology, which in the end might mean the planning application is rejected, requires significant redesign work or which could even bring the entire project to a halt. Even if the developer chooses to continue their project they will have to accept and pay for a full archaeology dig which is very costly. Understandably, developers are keen for a solution that provides faster assessments without losing accuracy. The position has become all the more pressing because in August 2020 the UK government announced their Build Back Better campaign. This updating of approach will reduce the time it takes to acquire planning permission from an average of 7 years down to 30 months. As a result, conventional archaeology assessments won’t be able to maintain their current high standards, which would lead to a reduction in their accuracy – something developers have raised significant concerns about in interviews. ArchAI covers this gap with instant, high accuracy assessment at an early stage. Founder Iris Kramer is using technology developed during her PhD in deep learning and previous degree in archaeology. She combines the two domains to deliver an innovative product that applies the power of AI to the construction industry to deliver rapid results and improved outcomes by automatically detecting archaeological sites on earth observation data. The product has already been successfully trialled with Historic Environment Scotland (HES) on the Isle of Arran, where they found hundreds of previously unknown archaeological sites and is now rapidly accelerating thanks to Iris’s successfully winning spaces in a number of programmes including Geovation (an accelerator programme run by the Ordnance Survey), the Royal Academy of Engineering Enterprise Fellowship scheme and culminating in her winning investment in this year’s Future Worlds Dragons’ Den event where ArchAI achieved a record valuation. Plans are underway to integrate ArchAI’s services into a web application where developers can select potential areas for development. ArchAI will then instantly provide archaeology predictions to help developers to quickly identify what areas to avoid and which will be most suited for development. This instant assessment can bring a huge impact to large infrastructure projects and Iris is already in discussions with a number of large construction firms and has already agreed to deliver a paid trail working on a major infrastructure project with one of the key players in the construction field and a government agency. ArchAI is seeking partnerships with large construction companies, particularly those involved in major infrastructure projects where she can quickly bring significant, scalable benefits. She is also considering the possibility of bringing on board a co-founder with either an established reputation and connections at board level within the construction industry or extensive experience in using deep learning and would be interested to hear from anyone who fits that profile. With instant, high accuracy assessments over large areas at the earliest planning stages, ArchAI is de-risking and saving time and money for the construction industry while also saving vital historical sites from unnecessary destruction.