By
Stuart Kerr, Technology Correspondent, LiveAIWire
Archaeology has always been a discipline of reconstruction:
working backward from fragments of material culture to the lives of people
who left no living testimony. Artificial intelligence does not change this
fundamental reality, but it is changing the speed, scale, and precision with
which fragments can be identified, analysed, and connected. The applications
range from the detection of buried sites from satellite imagery to the
reassembly of shattered ceramic vessels, the decipherment of damaged
manuscripts, and the digital reconstruction of structures lost to conflict or
time.
Seeing Through the Ground: Satellite and LiDAR Analysis
Machine learning systems trained to recognise the spectral or
topographic signatures of archaeological features can process satellite
imagery, aerial photography, and LiDAR data at regional and continental
scale. Research published in Nature Astronomy in 2022 documented the
discovery of hundreds of previously unknown Nazca geoglyphs in Peru using
machine learning analysis of aerial photography, more than doubling the
number of known figures in the region. In Honduras, LiDAR analysis revealed
the full extent of a lost Maya city whose surface remains had been known but
whose scale was only apparent from aerial survey data. The Antiquity
journal has published multiple studies demonstrating AI-assisted
remote sensing identifying things that human analysts would not find at all
within realistic resource constraints, not merely finding them more
slowly.
Reassembly and Reconstruction: The Puzzle-Solving
Machine
Computer vision systems trained on three-dimensional scan data can
identify matching edges, surfaces, and curvature profiles between fragments
in datasets of thousands of pieces, generating reassembly suggestions for
human experts to evaluate. The Antikythera Mechanism, the ancient Greek
astronomical computer recovered in fragmentary form in 1901, benefited from
AI-assisted analysis of X-ray and surface scan data published in 2021 that produced
the most detailed reconstruction yet proposed of its complete structure and
function, demonstrating that AI can extract interpretive insight from
existing data that human analysis had not exhausted.
What This Means for You
Digital reconstruction of heritage sites enables accessible
visualisations of places that are physically inaccessible, partially
destroyed, or located in regions that most visitors cannot reach. The
reconstruction of Palmyra’s Arch of Triumph, destroyed in 2015, using
AI-assisted photogrammetric analysis of pre-destruction photography became
both a scholarly record and a component of public heritage advocacy. Similar
projects are now underway for heritage sites damaged in Ukraine, Yemen, and
Afghanistan. As LiveAIWire has covered in analysis of AI
transforming traditional knowledge domains, AI is most valuable
when it augments deep domain expertise rather than attempting to replace it,
and archaeology is no exception.
Decipherment: Reading What Was Unreadable
The Vesuvius Challenge, launched in 2023, offered prizes for the
use of AI to read carbonised scrolls from Herculaneum preserved but rendered
illegible by the 79 CE eruption of Vesuvius. By the end of 2023, teams using
machine learning trained on CT scan data had successfully read substantial
portions of scroll text, including philosophical content inaccessible for
nearly two thousand years. The British
Library’s digital scholarship blog documents multiple projects
using machine learning for handwriting recognition, date estimation, and
identification of scribal hands across its digitised manuscript collection,
enabling research that would require decades of human expert analysis to approach
at conventional speed.
Repatriation, Ethics, and the Digital Double
Digital reconstruction of cultural heritage objects raises complex
questions for communities whose material culture was removed through colonial
acquisition. A high-resolution AI-generated digital model does not restore
the relationships between communities and their material heritage that
physical repatriation would. At the same time, digital access provides forms
of connection and scholarship that are genuinely valuable. As LiveAIWire has
examined in coverage of AI
and data equity questions, who benefits from AI-generated value
derived from particular communities’ knowledge and cultural production is a
consistent ethical challenge of the technology’s current
phase.
Preservation at Scale
Climate change, conflict, urban development, and the simple
passage of time are destroying archaeological sites at a rate that
conventional documentation methods cannot keep pace with. The International Council on
Monuments and Sites has identified AI-assisted heritage
documentation as a priority capability, recognising that the combination of
photogrammetry, LiDAR, multispectral imaging, and machine learning analysis
can capture material characteristics of a site in more detail than any
previous documentation method. The challenge is ensuring data is stored in
formats that remain accessible as technology evolves, and that analysis is
documented in ways that future scholars can critically
assess.
AI and the Management of Archaeological
Knowledge
Beyond individual site discoveries and object analyses, AI is
beginning to transform the management and accessibility of archaeological
knowledge at an institutional level. The digitisation of museum collections,
combined with AI-powered image recognition and metadata generation, is making
it possible to search visual archives at a scale and specificity that was
previously impractical. A researcher seeking parallels for a specific ceramic
form or architectural feature can now search across digitised collections
from dozens of institutions simultaneously, identifying comparanda that would
previously have required years of manual archive work.
Machine learning is also being applied to the integration of
heterogeneous datasets: combining excavation records, environmental data,
radiocarbon dates, archaeobotanical analyses, and isotopic results into
unified interpretive frameworks. These integration tasks are technically
demanding and have historically been bottlenecks in archaeological
publication, with significant time elapsing between data collection and
interpretive synthesis. AI tools that can help identify consistent patterns
across disparate data types and flag anomalies that warrant further
investigation have the potential to substantially accelerate this
process.
The open science movement in archaeology, which advocates for the
publication of primary datasets rather than only interpretive conclusions,
creates the large-scale, well-documented datasets that AI systems need to
learn from effectively. The combination of open data practices and machine
learning capability is creating a virtuous cycle in which more data enables
better models, and better models make more data worth collecting. The
long-term implications for the pace of archaeological knowledge production
are potentially significant, though the field’s characteristic caution about
interpretive claims will remain an appropriate counterbalance to the
enthusiasm that new technologies always generate.
Training the Next Generation of Archaeological
AI
The effectiveness of AI in archaeology depends fundamentally on
the quality and comprehensiveness of the training data available to machine
learning systems. Archaeological datasets present particular challenges: they
are heterogeneous in format and quality, they reflect the historical biases
of a discipline that has concentrated excavation resources in certain regions
and periods, and they are often held by institutions that have not yet made
their holdings fully digitally accessible. Building the datasets that would
enable more powerful archaeological AI requires coordinated effort across
institutions, national archaeological bodies, and the research
community.
Several initiatives are working toward this goal. The
Archaeological Data Service in the UK aggregates and preserves digital
archaeological datasets, and its holdings are increasingly being used as
training data for machine learning research. ARIADNE, the European
archaeological research infrastructure, is building interoperable data
standards and access systems that will enable AI researchers to train on
datasets spanning multiple national traditions and time periods. These
infrastructure investments are essential prerequisites for the next
generation of archaeological AI tools, and their development has received
significant support from European research funding
programmes.
The participation of archaeologists in the development of AI tools
for their discipline is equally important. Domain expertise is essential for
defining appropriate training datasets, evaluating model outputs critically,
and identifying the failure modes that purely technical evaluation would
miss. The most effective archaeological AI projects have been collaborations
between computer scientists and archaeologists from their inception, with
domain knowledge shaping technical choices throughout the development
process. This model of collaborative development, in which disciplinary
expertise guides rather than simply validates technical work, is the approach
most likely to produce tools that genuinely advance archaeological
knowledge.
About the Author
Stuart Kerr is the Technology Correspondent at LiveAIWire,
covering artificial intelligence across society, policy, and industry. About
LiveAIWire.