Why it matters
Without it, miscatalogued finds, missing illustrations, inconsistent captions, and fragmented reporting slow post-excavation assessment and archive deposition, creating backlogs and potential non-compliance with HER and CIfA standards.
From design to safety
Design
AI-assisted selection criteria setup
The system reads the written scheme of investigation, project brief, and HER requirements to define catalogue fields, plate types, and caption standards before field data collection begins, ensuring alignment with existing HER entries and context numbering conventions, with an auditable trail of versioning and stakeholder approvals.
- plate rack
- caption template sheet
- digital asset management server
Build
Data and asset integration
It ingests context sheets, finds records, and reference images, mapping them to planned plates and establishing initial illustration pipelines for post-excavation, including cross-references to Harris matrix and context plan layers, plus preservation of physical and digital provenance logs in the archive.
- metal filing cabinet
- photo lightbox
- caption template sheet
Operation
Live selection and captioning
The AI recommends plates and captions as finds are processed, cross-checks metadata accuracy, and aligns outputs with agreed plate counts and illustrative formats, flagging discrepancies to the finds supervisor for review, and tagging anomalies for curator scrutiny.
- photo lightbox
- plates binder
- digital asset management server
Maintenance
Asset quality and metadata curation
Ongoing calibration of image quality checks, caption terminology, and provenance tagging ensures long-term searchability and compliance with archiving standards, including periodic audits of plate types against project specs and updated thesauri for terminology.
- captioning workstation
- plate rack
- digital asset management server
Safety
Integrity and archive readiness
The system flags incomplete records, potential copyright concerns, and misaligned captions, triggering QA by the archivist before archive deposition, with final sign-off by the HER officer on metadata integrity and deposit readiness.
- metal filing cabinet
- digital asset management server
- caption template sheet
Part by part
- 1
metal filing cabinet
- Watch for
- misfiled or lost finds entries with incorrect context numbers or catalog codes, risking misassociation of finds within a context or feature
- AI action
- auto-flag mismatches and route to cataloguer for correction ensuring traceability to context sheets and Harris matrix
- 2
digital asset management server
- Watch for
- duplicate asset IDs or missing provenance fields in metadata
- AI action
- auto-deduplicate and populate missing fields from linked context sheets with revision logs
- 3
photo lightbox
- Watch for
- photos lacking scale, lighting, or correct orientation
- AI action
- prompt photographer to reshoot or auto-correct metadata in system with prioritised plate types
- 4
plates binder
- Watch for
- plates without plate number or plate type
- AI action
- assign plate IDs and generate draft captions template with cross-links to context
- 5
captioning workstation
- Watch for
- inconsistent terminology across captions
- AI action
- standardize vocabulary and suggest uniform phrasing for captions with HER-compliant language
- 6
caption template sheet
- Watch for
- missing required caption fields (plate number, object, material, interpretation)
- AI action
- auto-fill with prompts and flag gaps for curator approval ensuring cross-reference to plates and context
Worked example
Site 12A post-excavation finds and illustration workflow
Illustrative scenario with sample data.
A 4 ha development-led site requires coherent finds catalogue and illustration pack for the post-excavation assessment and HER submission.
Signal in
context sheets: Ctx-12A-001 to Ctx-12A-050; finds IDs F-001 to F-420; plate types: planchette, section plate, illustrative plate; metadata: date, findspot, object type, material; sample: 20 environmental plates.
System out
draft catalogue entries for 420 finds; 150 plates allocated; captions drafted for 120 plates; report-ready illustration pack linked to HER submission.
Human sign-off: Post-excavation manager and HER officer approve the final illustration set and archive deposition plan.
How to roll it out
Phase 1
Weeks 1 to 4
Baseline setup of asset management, plate templates, and captioning rules in the project library.
Phase 2
Weeks 5 to 8
Data ingestion pipelines and initial automated checks on finds and images integrated with the WSI and HER requirements.
Phase 3
Weeks 9 to 12
Live production of catalogue and plates with QA reviews by archivist and reporting officer.
Where people stay in charge
- Final captions, plate selections, and attribution decisions must be signed off by the archivist or HER officer to assure provenance.
- All asset outputs must be traceable to linked context sheets and supported by an auditable change log.
- Copyright, reuse rights, and archival deposit approvals require sign-off from the curator and client heritage consultant before public dissemination.
Questions owners and crews ask
What is AI doing in post-excavation finds and illustration selection?
AI reads finds metadata, aligns with plate types, selects applicable illustrations, and drafts captions. Final approval rests with the archivist and HER officer.
How does AI handle inconsistent captions across reports?
It analyzes terminology, flags inconsistencies, and suggests standardized wording for curator and reporting officer approval.
Can the system work with legacy archive formats?
Yes, it maps legacy records to current schema, preserves provenance, and routes any ambiguities to the archivist for resolution.
Who signs off the final illustrated post-excavation report?
The archivist with the client’s heritage consultant signs off the final archive deposition and reporting package.