NAVIGATING A NEW DIGITAL INTERFACE: USING AUTOMATED IMAGE RECOGNITION TO IDENTIFY POTTERY IN THE ARCHAIDE PROJECT

Archaeological Automatic Interpretation and Documentation of cEramic (ArchAIDE) is a H2020 funded project (2016-2019) developing digital tools to support archaeologists in recognising and classifying pottery. ArchAIDE is not designed to replace the knowledge of pottery specialists, but seeks to complement by speeding time consuming tasks, provide support for non-specialists, help students learn more about pottery recognition, and aid in the collection of metadata needed to describe the pottery. ArchAIDE is developing a tablet-based mobile app which relies upon image recognition and uses deep learning to narrow and suggest possible matches. While ArchAIDE has been careful to design a tool that allows classification decisions to be made by users at key points in the recording workflow, the app uses digital tools and methods for a significant tasks that were previously carried out using analogue methods. This paper will explore how users are engaging differently with the archaeology when using a digital workflow for identifying, classifying and recording pottery, as observed by the ArchAIDE project partners in early testing. This will include issues around using digitised comparative collections rather than paper catalogues, using the app to identify pottery while still in the field-rather than during post-excavation, how users might ‘see’ pottery differently through a digital rather than analogue analysis, and whether pottery identification using a digital interface changes knowledge transmission and learning processes. While the purpose of the ArchAIDE project is to make pottery identification faster and easier, this paper will pause to reflect and critically engage with moving to a digital workflow, and how this may influence how archaeological knowledge is produced and understood.