Contemporary archaeological survey practices vary substantially between individual practitioners, different regional traditions, ascribed intradisciplinary labels, and disciplinary shifts across time. These varied approaches need to be untangled and quantified before they can be integrated with automated methods in a way that is explainable, accountable and ethical.
︎
Contemporary archaeological survey practices also have long histories and the introduction of technological tools isn’t new. Both ground-level and aerial survey practices have been augmented by digital processes providing varying levels of input by machine since the 1970s. This has affected both survey practices and their products, as well as creating an extending suite of data collection, analysis and management methods.
ML and CV are fundamentally new tools in archaeology and represent a meeting point between these different approaches to survey practice. This project will be examining these interconnected histories, technological tools and routines of practice by taking a detailed look at workflows for archaeological survey. This proactive approach to understanding the human biases which can be implicitly or uncritically included within survey practice is essential groundwork for the ethical integration of automation with archaeological survey.
︎
This project is based within the Rapid Archaeological Mapping Project at Historic Environment Scotland. Drawing on participatory design practices, the project will be working in collaboration with this team to assess, design and test workflows. The project will also learn from the workflows and expertise of other projects, practitioners and colleagues working across Europe.
If you’d like to know more, or to get in touch to talk about the project, please click ︎ for contact info.
Automation in the practice of archaeological survey(1)—Integrating Machine Learning(2), Computer Vision(3), People and Practice(4).