Why it matters
Without it, misreads of historic evidence can lead to unexpected discoveries, delays in planning conditions, and unaddressed HER constraints that slow construction and complicate compliance.
From design to safety
Design
AI-guided map regression inputs
AI reads historic maps, tithe maps, OS editions, aerial archives, and HER records to identify zones of potential heritage significance for desk-based assessment, ensuring a comprehensive starting framework for written schemes and trench design, while documenting source confidence and cross-referencing archive descriptions.
- historic maps
- tithe maps
- Ordnance Survey editions
- aerial archives
Build
Data-informed trench planning
AI synthesizes map-derived constraints with site survey data, GIS layers, and ground-truth checks to adjust trench layouts and integration with the written scheme of investigation workflows in real time, logging changes to the Harris matrix and context intervals.
- Ordnance Survey editions
- aerial archives
- HER records
- desk-based assessment
Operation
Context-driven record updates
During excavation, AI aligns context sheets and digital plans with map-based cues to refine target features, ensure watching briefs, and support ongoing client reporting and auditing, while updating the archive search path.
- historic maps
- HER records
- desk-based assessment
- tithe maps
Maintenance
Archive readiness and re-evaluation
AI tracks provenance of map-derived data, flags data gaps for post-excavation assessment, schedules archive deposition, HER updates, and ensures long-term accessibility for researchers, including data migrations and metadata standardization.
- desk-based assessment
- aerial archives
- HER records
- Ordnance Survey editions
Safety
Regulatory and safety alignment
AI reviews mapping-derived constraints for buried services and planning conditions to support safe trenching, risk assessment, and compliance oversight by the planning authority, with incident logging and corrective actions.
- historic maps
- Ordnance Survey editions
- aerial archives
- tithe maps
Part by part
- 1
historic maps
- Watch for
- Map features aligned with older cartography that diverge from current ground truth beyond expected tolerances.
- AI action
- Read cartographic offsets and trigger field verification workflows.
- 2
HER records
- Watch for
- Conflicting or updated HER entries indicating new allocations or COIs near planned trenches.
- AI action
- Cross-check against current permits and trigger HER consultation.
- 3
tithe maps
- Watch for
- Inconsistent parish boundaries or scale errors affecting site relevance.
- AI action
- Flag boundary discrepancies and prompt desk-based re-check.
- 4
Ordnance Survey editions
- Watch for
- Edition gaps or misalignments with modern grid causing misplacement risk.
- AI action
- Detect edition alignment issues and propose re-geo-referencing.
- 5
aerial archives
- Watch for
- Shaded relief or flight-era biases masking features of interest.
- AI action
- Highlight low-res features for higher-resolution follow-up.
- 6
desk-based assessment
- Watch for
- Inadequate bibliography or missing cross-references to primary sources.
- AI action
- Suggest additional sources and generate a draft assessment outline.
Worked example
Cumulative map regression supports desk-based assessment on a 2.3 ha development
Illustrative scenario with sample data.
At a 2.3 ha development, an early desk-based assessment uses AI to collate historic maps and HER records for potential constraints during trench design.
Signal in
site=Southfield Wood, area=2.3 ha; maps=[historic_maps:1780, 1880_OS:1:1250, tithe_map:1830], HER=[CMS:1274, HERID:5012], aerial_arch=[aerial_1950, aerial_1970]; data_fields=[grid_ref, scale, date, source, reliability]
System out
AI-generated desk-based assessment draft highlighting probable zones of archaeological interest and recommended HER consultations; report_id=DBA-2203-HA; priority=Medium.
Human sign-off: Project manager reviews the draft, authorizes field verification, and coordinates HER consultation with the planning archaeologist.
How to roll it out
Phase 1
Weeks 1 to 4: Baseline
Establish data ingestion pipelines, asset indexing, and initial desk-based workflow with map sources and HER integration.
Phase 2
Weeks 5 to 8: Validation
Cross-check AI outputs against known investigations, calibrate thresholds, and refine risk flags for trench planning.
Phase 3
Weeks 9 to 12: Deployment
Integrate AI outputs into live design and monitoring; prepare audit-ready records and archive deposition plan.
Where people stay in charge
- All regulatory decisions remain signed off by the qualified person in charge of planning conditions and HER compliance.
- Data handling and archival actions must be approved by the client's heritage consultant or local planning authority where required.
- Safety interpretations must be verified by a site supervisor before field execution.
Questions owners and crews ask
What sources does the AI use for desk-based assessment?
The AI reads historic maps, tithe maps, Ordnance Survey editions, aerial archives, and HER records, cross-referencing with current site data to flag zones of potential significance and to support the desk-based assessment.
How does AI help with trench design?
It identifies heritage constraints from map-derived cues, integrates them into the written scheme of investigation, and informs trench layouts, ensuring alignment with planning conditions and camera-ready context recording.
Who approves AI-generated outputs?
The project manager reviews, then the planning archaeologist or HER officer signs off where required before fieldwork proceeds, with archival copies stored in the HER system.
What safeguards ensure data quality and compliance?
All sources are logged with provenance, scale, date, and reliability; outputs are auditable and aligned to CIfA standards, HER guidance, and consent conditions, with versioned records.