Archaeology · AI · Data Science · Computational Modelling

I am an archaeologist and AI/data scientist interested in building computational systems that make archaeological processes measurable, modelled and testable.

I run Industrial Automata, an independent technical advisory and research practice in Ireland.

My work sits at the intersection of archaeology, software engineering, artificial intelligence and scientific modelling. I am particularly interested in moving from archaeological traces to explicit models of the processes that produced them.

A central idea in my work is to treat archaeological evidence as observable signals of physical processes, rather than beginning with cultural or typological categories.

This leads to a computational research loop:

trace → recognise structure → generate candidate process → forward-simulate → predict trace → compare → revise

The aim is to connect archaeological observation with mechanisms that can be implemented, simulated and tested.

Published: D. Foley and E. Furey, "From Geospatial Patterns to Ancient Signals: A Signal-Based Framework for Archaeological Machine Learning," 2025 35th Irish Signals and Systems Conference (ISSC), IEEE. doi:10.1109/ISSC67739.2025.11291309

Current research interests include:

• archaeological signal modelling
• process-based archaeology
• geospatial and temporal data
• lifecycle and motion modelling
• AI and machine learning
• robotic simulation
• computational approaches to technology and behaviour

Each project applies the loop above to a different kind of evidence. Planned work is tracked as issues on the public Research programme board.

• lithic-motion-traces: recovering the strikes that shaped a stone tool from its 3D scan, by fitting a forward scar model to Aurignacian lithics from Grotta di Castelcivita (c. 41,000 years old).
• pu_ratchet_analysis: procedural-unit acquisition across 3.3 million years of the hominin lithic record. Reproduces from the public dataset, checked automatically on every change.
• streedagh-armada: three Spanish Armada wrecks at Streedagh, Co. Sligo (1588), and Captain Cuéllar's account, read as a signal field. Every claim is tagged as observed, derived, inferred or hypothesis.
• round-tower-acoustics: a boundary element model of how a voice in an Irish round tower's top chamber reaches people gathered outside.
• technosignature-survey: TESS light curves surveyed the way a site is: record every dip first, then explain it away. Began as an MSc exercise in 2023.

A machine-readable framework for representing archaeological observations as structured signals.

The system separates SENSED observations from INFERRED interpretations and uses measurable physical properties rather than typological labels as its fundamental data layer.

This provides a foundation for archaeological data that can be queried, transformed, simulated and ultimately used for machine learning.

A geometric representation of an object's history.

Instead of treating an archaeological object simply as something located at a point in space, the lifecycle vector represents its movement through spatial and temporal states—from its present condition towards its earlier states and processes.

The same abstraction can operate across very different timescales, from archaeological processes over millennia to robotic processes occurring over seconds.

A broader archaeological signal framework for converting spatial, temporal and observable archaeological evidence into computational representations.

The goal is to make archaeological datasets suitable for quantitative analysis without requiring cultural interpretation to be embedded in the underlying observation model.

I am developing a process-based model of technology in which technological behaviour is represented through interactions between an organism, its environment, motion and material.

This work is intended to connect archaeological traces with forward models and, eventually, robotic experiments.

An open-source Python point-of-sale system developed for a real bookshop.

Built with Python and Textual, Colophon is an example of my broader approach to software: build systems that solve actual problems, deploy them in the real world, observe what happens, and improve them.

View Colophon →

MSc — Artificial Intelligence & Big Data, Atlantic Technological University, Letterkenny (2024)

BA (Combined Honours) — Archaeology & Information Technology, University of Galway (2002)

~15 years of archaeological fieldwork, including running my own consultancy, Archaeological Investigations

My background combines archaeological field experience with formal training in computer science, artificial intelligence, programming and data analysis.

I am interested in the space between disciplines: taking problems that originate in the physical world and translating them into computational systems that can be measured, modelled and tested.

I am particularly interested in the eventual convergence of:

• archaeological evidence
• computational representations
• forward simulation
• machine learning
• robotics

The longer-term objective is a scientific framework in which archaeological traces can be used not only to describe the past, but to construct and test models of the processes that produced them.

Building computational systems for archaeology.