Why it matters
Without AI, tear repair decisions risk inconsistency in materials and sequencing, misaligned infills, and failed reversibility goals, leading to irreversible damage, missed loan deadlines, and disputed insurance claims or damaged client trust.
From design to safety
Design
AI guides proposal and materials
The AI reads historic repairs, paper type, tear geometry, and anticipated loss extent to support treatment proposals and material selection.
- Japanese tissue
- wheat starch paste
- infills
Build
AI informs bench fabrication
The plan translates into bench instructions, including sheet-cutting guides, paste batch records, and leafcasting mold configurations aligned with documented inlays.
- leafcasting
- infills
- tear mending
Operation
AI monitors ongoing repairs
During treatment, the system tracks humidity, reversibility tests, color matching, and adhesion tests while updating the condition record and treatment proposal.
- Japanese tissue
- wheat starch paste
- leafcasting
Maintenance
AI supports archival records
Post-treatment, the AI logs outcomes, stores before-and-after scans, tracks material usage and batch traceability, and prompts follow-up condition checks for long-term stability.
- infills
- leafcasting
- tear mending
Safety
AI aids risk and compliance
The AI flags COSHH considerations, material hazards, and ethical approvals, ensuring sign-off by the accredited conservator and documentation for PACR compliance.
- Japanese tissue
- wheat starch paste
- humidification
Part by part
- 1
Japanese tissue
- Watch for
- Rapidly changing tear edge curl, weakness at hinge zones, uneven porosity risks weakening repairs.
- AI action
- Detect tear edge geometry and recommend tissue size, orientation, and alignment with surrounding fibers.
- 2
wheat starch paste
- Watch for
- Adhesive penetration beyond repair area, yellowing, embrittlement under humidity shifts.
- AI action
- Predict adhesive performance, propose optimal setting time, and specify moisture-control steps during pressing.
- 3
infills
- Watch for
- Infill shrinkage mismatch, color/texture separation from surrounding paper.
- AI action
- Suggest infill material density, pigment matching, and precise placement coordinates with edge blending guidelines.
- 4
leafcasting
- Watch for
- Uneven casting thickness, hydration shadows, edge bleed.
- AI action
- Forecast cast uniformity and drying profile, provide mold orientation and support sequence.
- 5
tear mending
- Watch for
- Inconsistent stitch length, visible repair traces, differential stiffness.
- AI action
- Recommend repair sequence, stitch spacing, stabilization intervals, and repeatability checks.
- 6
humidification
- Watch for
- Climatic fluctuations impacting cellulose integrity, rebound risk after humidification.
- AI action
- Monitor chamber humidity and trigger stabilization actions before proceeding to next steps.
Worked example
Tear and loss plan for a 19th-century book leaf
Illustrative scenario with sample data.
A conservator seeks a reproducible repair plan for a brittle leaf with multiple losses, ensuring reversibility and archival quality.
Signal in
paperType: rag; tearCount: 3; lossGeometries: irregular; humidityTarget: 45% RH; lighting: 500 lux; sampleIDs: L1-L3; binder: wheat starch paste; tissue: Japanese tissue; desired finish: matte
System out
Treatment proposal generated: tissue choice, leafcasting approach, adhesive plan, and documentation template; work order created; condition report template updated.
Human sign-off: Accredited conservator reviews AI-generated plan, confirms materials, signs off on treatment proposal, and approves the documentation scheme for PACR.
How to roll it out
Phase 1
Weeks 1 to 2
Baseline data capture, model calibration with existing records and standard treatments, plus integration with condition-report templates.
Phase 2
Weeks 3 to 6
Prototype planning outputs linked to bench workflows, first synthetic tests, and documentation templates for treatment reports and archives.
Phase 3
Weeks 7 to 12
Full integration with condition reporting, loan documentation, and archival records workflow, including PACR-ready files and provenance logs.
Where people stay in charge
- All final material selections and treatment sequencing must be approved by the accredited conservator.
- All reversible methods must be prioritized; every non-reversible choice requires explicit rationale and curator sign-off.
- Documentation and provenance records must be retained in PACR-compliant archives with clear editors and dates.
Questions owners and crews ask
What can AI help with in tear and loss planning for paper conservation?
AI assists treatment planning by analyzing tear geometry, loss area, and historical repairs to propose materials, placement, and workflow while maintaining traceable records for future reference.
How does AI affect documentation and records?
AI generates structured treatment proposals, work orders, and condition reports that are archived in the studio’s records system with full provenance for future reference and audit.
What safety considerations are handled by AI?
AI flags adhesive hazards, humidity control needs, and reversible options; sign-off remains with the accredited conservator and COSHH compliance checks are recorded and filed.
Who approves the AI-generated plan?
The accredited conservator reviews and signs off on the treatment proposal, with curator and client input when required, ensuring ethical compliance and PACR readiness.