Why it matters
Without AI, irreversible treatment risks, solvent incidents, missed insurance or loan deadlines, and disputes over material provenance, reducing client trust and complicating compliance with PACR accreditation and COSHH.
From design to safety
Design
AI informs treatment planning proposals
AI analyzes initial technical examination data, material lists, and regulatory constraints to draft a treatment proposal with reversible methods and carefully chosen materials, including consolidants, cleaners, and inpaint formulations aligned with minimal intervention. It integrates pigment layer thickness from SEM, micro-CT fracture risk scores, and enamel strength data to constrain solvent choices and line up with PACR ethics.
- Hyperspectral imaging camera
- Portable X-ray fluorescence (pXRF) analyzer
- Infrared thermography camera
Build
Systems set for safety and data capture
The studio specifications include imaging mounts, fume extraction connections, and data capture templates for later audit of condition and processes, ensuring traceable chain-of-custody, supplier verifications, and documented bench protocols. It enforces reversible adhesion tests and records consolidant batch IDs with lot-level documentation and failure-mode logging.
- Hyperspectral imaging camera
- IR thermography camera
- Digital microscope with camera
Operation
Ongoing monitoring and decision support
AI processes serial condition records, imaging data, and regulatory updates to flag anomalies and propose adjustments to the treatment plan, sequencing, sampling, and consolidation strategy based on cumulative evidence. It flags colorimetric shifts, gloss changes, and microcrack progression with timestamped audit entries.
- Portable XRF analyzer
- Digital microscope
- SEM-EDS sample stage
Maintenance
Calibration and data hygiene routines
Regular calibration of imaging and spectroscopy devices, and validation of data streams against reference materials for consistent decision support, with versioned treatment templates and automated audit trails. It logs reagent shelf-life, storage temperature, and staff access events for traceability.
- Infrared thermography camera
- Hyperspectral imaging camera
- pXRF analyzer
Safety
Ethical and regulatory decision governance
The system enforces reversibility, records-keeping, and COSHH compliance, with final approvals by the accredited conservator and regulatory sign-off bodies, and routine ethics reviews documented in the PACR archive. It maintains hazard assessments, ventilation rates, and PPE usage records.
- Fume extraction hood
- COSHH monitoring system
- Regulatory compliance dashboard
Part by part
- 1
Hyperspectral imaging camera
- Watch for
- Drift in spectral signatures indicating calibration lapse or surface contamination.
- AI action
- AI detects drift and prompts recalibration or retake of spectral data.
- 2
Portable X-ray fluorescence (pXRF) analyzer
- Watch for
- Shifts in elemental composition suggesting degradation or contamination layers.
- AI action
- AI flags anomalies and recommends sample locations.
- 3
Infrared thermography camera
- Watch for
- Thermal gradient irregularities indicating voids or delamination.
- AI action
- AI highlights hotspots and guides where to inspect.
- 4
Digital microscope with camera
- Watch for
- Microcracking or microfissures in enamel and tortoiseshell interfaces.
- AI action
- AI annotates defects and proposes conservative cleaning tests.
- 5
Fume extraction hood
- Watch for
- Exceedance of airborne solvent exposure limits.
- AI action
- AI triggers ventilation adjustment and exposure logging.
- 6
Regulatory compliance dashboard
- Watch for
- Non-conformance with PACR, COSHH, or local regulations.
- AI action
- AI issues compliance alerts and flags required actions.
Worked example
AI-assisted assessment of enamel and ivory composite join stability
Illustrative scenario with sample data.
A private studio evaluates a 19th-century enamel plaque with ivory inlays and a composite join, seeking reversible stabilization and regulatory alignment.
Signal in
material_id: enamel-ivory-combo; surface_condition: crazing; spectroscopy: pXRF_readout=Si:22, Ca:18; imaging: hyperspectral=band_edges; regulatory_status: PACR_pending
System out
treatment_proposal: reversible consolidant selection; monitoring plan; regulatory documentation initiated; recommended follow-up imaging schedule; owner notified of compliance plan.
Human sign-off: Head conservator signs off on the proposal, with curator review for aesthetic considerations.
How to roll it out
Phase 1
Weeks 1 to 2
Baseline data capture of condition, imaging, and material records; establish imaging workflow.
Phase 2
Weeks 3 to 6
Prototype AI alert rules and calibration checks; integrate into treatment planning templates.
Phase 3
Weeks 7 to 12
Full operational deployment with audit trails, client-facing reports, and compliance dashboards.
Where people stay in charge
- All material decisions must be approved by the accredited conservator after review of imaging data and treatment options.
- The AI may suggest only reversible methods and must not override ethical constraints or statutes.
- Final sign-off on treatment and records rests with the PACR-accredited head of conservation.
Questions owners and crews ask
What data does the AI read for enamel and ivory assessment?
The AI reads imaging data, spectroscopy results, surface condition notes, and regulatory flags to propose reversible treatments and document compliance.
Who approves AI-generated treatment proposals?
The accredited conservator reviews AI outputs, with curator input for aesthetic implications before client approval.
How are records kept for future provenance?
All imaging, spectroscopy, and treatment decisions are stored in a versioned project archive with audit trails.
What safety checks accompany AI recommendations?
COSHH, ventilation status, and ethics considerations are validated by the conservator before implementation.