Why it matters
Without this, miscatalogued images, missing context numbers, and inconsistent metadata lead to delays in post-excavation analysis, compromised archives, and risky re-interpretation of finds.
From design to safety
Design
AI aids site photography planning
AI reviews the written scheme of investigation and proposed photography angles, scales, and context numbers to set standardized metadata schemas and capture plans. It aligns with trench plans and pre-validates north arrow placement. It simulates coverage to guide field teams.
- digital camera
- photo register
- context number placards
- north arrows
Build
Capture data with standardized checks
During trench photography, AI flags missing scales, misoriented north arrows, or incomplete context markers to ensure consistent field records. It cross-references plan sheets and the Harris matrix, and prompts re-shoots. It logs metadata and maintains the photo register.
- digital camera
- north arrows
- scale bars
- photo register
Operation
Ongoing cataloguing and crosslinking
AI tags new photographs with context IDs, links them to plans, and updates the photo registers and metadata cache for rapid access. It cross-links to context sheets and photogrammetry outputs, and maintains archive integrity with GUIDs.
- photo register
- context number placards
- scale bars
- metadata schema
Maintenance
Data integrity and archive prep
AI audits metadata fields, verifies context numbers against sheets, and prepares batches for post-excavation assessment. It validates file integrity via checksums, reviews access logs, and schedules deposition-ready exports.
- context number placards
- photo register
- metadata schema
- GPS/data logger
Safety
Audit trails and compliance checks
AI maintains immutable records of tagging decisions, flags anomalies, and ensures adherence to CIfA standards for archive deposition. It reviews image lineage, flags misfiled records, and supports controlled access for sign-offs.
- digital camera
- north arrows
- photo register
- metadata schema
Part by part
- 1
Digital camera
- Watch for
- Missing or misaligned capture, wrong exposure, or absent metadata identifiers.
- AI action
- Detect missing fields and prompt re-shot or metadata fill-in.
- 2
Photo register
- Watch for
- Unmatched photo IDs, duplicate entries, or gaps in sequence.
- AI action
- Flag gaps and auto-link new photos to context sheets.
- 3
Context number placards
- Watch for
- Inconsistent or illegible context numbers and placard placement.
- AI action
- Validate against context sheets and trigger retake if needed.
- 4
North arrows
- Watch for
- Incorrect orientation or missing directional markers.
- AI action
- Compare to plans and flag orientation errors.
- 5
Scale bars
- Watch for
- Scale absent or inappropriate for the shot scale.
- AI action
- Recommend retake with proper scale reference.
- 6
GPS/data logger
- Watch for
- Missing geolocation or timestamp data in image metadata.
- AI action
- Impute or verify via field records and log integrity check.
Worked example
Archive-ready photo set for new development trench
Illustrative scenario with sample data.
A 2.8 ha development trench yields hundreds of photographs during design and build, with multiple contexts requiring consistent tagging for post-excavation and HER submission.
Signal in
site: Riverside Lane, trench 4, date: 2026-07-14, sensor: DSLR, lens: 18-55, lensFlag: 35mm, fields: context_id, north_arrow, scale, metadata, file_id, time_stamp, GPS, camera_model, exposure, white_balance, focal_length, capture_device, shot_id, author_id.
System out
AI-generated catalog with 1) standardized file names; 2) linked context IDs; 3) verified north arrow orientation; 4) attached scale bars; 5) complete metadata block; 6) archive-ready HTML index; 7) audit log entry and archive deposition ticket.
Human sign-off: Post-excavation manager approves final catalog after cross-check with context sheets and HER requirements.
How to roll it out
Phase 1
Weeks 1 to 4
Baseline data model, schema alignment with written scheme of investigation, and pilot photo register integration.
Phase 2
Weeks 5 to 8
Full-field deployment with automated tagging, cross-linking, and metadata validation workflows.
Phase 3
Weeks 9 to 12
Archive deposition workflows, quality assurance, and handover to client heritage team.
Where people stay in charge
- Decisions on archival deposition and metadata schema must be signed off by the project manager or a designated archivist, ensuring compliance with HER standards.
- All automation decisions affecting context linkage and archive integrity must be reviewed by a senior archaeologist before final deposition.
- Any system-suggested changes to context numbering or plan links require approval from the client’s heritage consultant before implementation.
Questions owners and crews ask
How will AI handle context numbers and context sheets during cataloguing?
AI reads context numbers from placards and cross-references them with context sheets to ensure correct links. Mismatches trigger a review prompt.
What metadata fields are essential for post-excavation assessment?
Essential fields include file_id, time_stamp, GPS coordinates, orientation, scale, context_id, and a complete metadata block for archive deposition.
How does AI ensure consistent photo register entries across trenches?
AI checks entries against trench IDs and plan sheets, flags gaps, and auto-links photos to contexts for consistency.
Who signs off on the final catalogued archive?
The project manager, with oversight from the client’s heritage consultant and the archaeologist in charge, signs off for compliance and HER readiness.