Why it matters
Without AI planning, fragment reassembly risks irreversible misfits, excessive handling, and disputes over treatment, leading to loan delays, insurance claim issues, and eroded client trust.
From design to safety
Design
AI-guided reassembly planning
The AI analyzes shard geometry, glaze signatures, provenance notes, 3D laser scans, infrared reflectography, and UV fluorescence to propose an initial fragment graph, alignment strategy, material choice, and tolerance bands before handling.
- shapes and curvature data
- fragment matching algorithm
- gap-fill strategy
Build
Prototype assembly workflow
The system schedules benchtop tests, pre-adhesive compatibility checks, jig design, mill-cut alignment fixtures, and 3D-printed supports using surface scans, substrate data, and archive references to produce a vetted treatment proposal.
- sherds
- fragment matching
- adhesives
Operation
Real-time alignment and documentation
During handling, AI tracks fit progress against the proposed graph, records precise alignments via high-resolution imaging and 3D coordinate logs, and updates condition and treatment records in real time.
- gap fills
- alignment
- adhesives
Maintenance
Model calibration and data hygiene
Periodic recalibration uses new fragment data, observed damages, and updated conservation records to maintain accuracy, track material behavior, and verify long-term compatibility of adhesives, gap-fill methods, and imaging signatures.
- glass disease
- shapes and curvature data
- fragment matching
Safety
Provenance and safety controls
The system flags ethical concerns, ensures reversibility, documents COSHH-compliant handling of adhesives and solvents, and requires sign-off by the senior conservator before irreversible work proceeds.
- sherds
- gloss and glaze data
- adhesives
Part by part
- 1
sherds
- Watch for
- risk of micro-fractures accumulating under pressure during handling and transport
- AI action
- detects high-stress zones and suggests reduced torque or alternative handling sequences during testing
- 2
fragment matching
- Watch for
- false positives from repetitive motifs or heavily worn edges creating mismatches
- AI action
- proposes best-fit candidates and ranks by predicted stability with confidence scores
- 3
adhesives
- Watch for
- chemical incompatibility with original materials or previous consolidants
- AI action
- selects adhesive type and curing conditions with documented reversibility and traceable provenance
- 4
gap fills
- Watch for
- shrinkage and color mismatch over time or differential aging
- AI action
- recommends fill material and deposition method with color-matching guidance and aging compatibility
- 5
glass disease
- Watch for
- microcrystalline changes indicating deterioration or demineralization patterns
- AI action
- triggers condition-monitoring protocol and preventive conservation plan with event logs
- 6
alignment
- Watch for
- misalignment risks causing stress concentration in reassembled object
- AI action
- provides alignment sequence, 3D coordinate verification, and final fit validation
Worked example
Planning a fragment reassembly for a 17th-century maiolica cup
Illustrative scenario with sample data.
A private studio plans to reassemble ceramic fragments with suspected glaze loss and visible cracks, aiming for faithful glaze continuity and structural integrity.
Signal in
fragment_id, shard_id, glaze_type, curvature_metric, color_signature, provenance_label, expected sequence, adhesive_type, area_measured_mm, initial bonding test results, substrate status, infrared_trace_profile, UV_fluorescence_map, laser_scan_data, journal_entry_id, access_control_level
System out
treatment_proposal_id: REA-2026-07, graph: G, suggested gap-fill materials, alignment jig plan with tolerances, material compatibility notes, cost estimate, risk register, alignment report, provenance continuity stamp, material safety data sheets linked
Human sign-off: Head of Conservation reviews AI-generated proposal, validates reversibility, and signs off on materials, handling plan, monitoring schedule, and archive integration; client is notified with formal condition report updates.
How to roll it out
Phase 1
Weeks 1 to 4
Baseline data capture, imaging protocol, and AI model validation using a representative fragment set with provenance records.
Phase 2
Weeks 5 to 12
Pilot reassembly on select fragments with adjoint documentation integrated into condition reports and treatment proposals.
Phase 3
Weeks 13 onward
Full deployment with ongoing data logging, audits, and formal staff training on workflow and safety protocols.
Where people stay in charge
- Decisions about treatment proposals, reversibility, and material choices must be signed off by a senior conservator or the head of conservation.
- The AI may propose matches and alignment but cannot approve the final arrangement without human peer review and ethics check.
- All owner and provenance data must be handled in accordance with professional guidelines and stored in the conservation records.
Questions owners and crews ask
What data does the AI use to plan fragment reassembly?
The AI reads fragment geometry, glaze and surface texture data, provenance notes, historical records, infrared trace data, UV maps, and measured curvature to propose matches and placement, with audit trails retained for each decision.
Who approves the AI's reassembly plan?
A senior conservator reviews the AI-generated plan, verifies reversibility and material compatibility, and signs off before any treatment proceeds.
How are gaps and fills decided?
Fill materials are chosen for compatibility, reversibility, and color match, with the plan documenting deposition sequences, containment, aging expectations, and long-term performance.
How is provenance protected in AI planning?
All provenance data are captured in condition and treatment records, with access controls, version history, and audit logs maintained.
Sources and further reading
- HSE: COSHHGuidance on controlling hazardous substances such as solvents and adhesives, with assessment requirements.
- ICOM Code of Ethics for MuseumsEthical standards for conservation and restoration of museum collections, applied to treatment decisions.
- Icon: ResourcesPractice and ethics guidance for conservators on decision-making, treatment responsibility, and professional standards.
- ICOM-CC Working GroupsSpecialist working groups covering metals and modern materials, with research on conservation of these materials.
- Getty Conservation InstituteConservation science research on imaging and examination techniques for objects and materials.
- Library of Congress: Preservation SciencePreservation research on materials analysis, spectral imaging, and testing methods for collection objects.
- ICOM Committee for ConservationTriennial conference publications on conservation of ceramics, glass, and other museum objects.
- ICCROM: Preventive ConservationAgents of deterioration and risk management methods for preventive conservation of heritage objects.
- Collections Trust: SpectrumCollections management standard covering condition checking and conservation procedure records.
- American Institute for Conservation: Collections CarePreventive care guidance on safe handling, storage, and display of collection objects.
- US National Park Service: Conserve O GramsTechnical notes on object care, handling damage, and related hazards for museum collections.
- American Institute for Conservation: Find a ConservatorDirectory of Professional Associate and Fellow conservators, searchable by specialism for choosing a conservator.