Why it matters
Without AI guidance, misselection risks irreversible treatment damage, solvent incidents, delayed loan deadlines, disputed insurance claims, and client dissatisfaction.
From design to safety
Design
AI informs treatment proposals
The AI analyzes data from condition reports, prior treatments, material compatibility, aging behavior, and expected solvent interactions to propose adhesive or consolidant options with defined reversibility profiles, saturation limits, aging trajectories, and long-term stability scenarios; it generates a literature-linked decision memo bridging data sources to material choices and rationale, including explicit sturgeon glue suitability notes and BEVA 371 compatibility matrices.
- Paraloid B-72
- BEVA 371
- sturgeon glue
Build
AI-guided material setup
AI-backed decision logs capture chosen adhesive/consolidant, supplier batch codes, storage conditions, and calibrated instrument readings; it prescribes bench-test matrices, test-patch locations, solvent test sequences, and interface data for controlled demonstrations in the bench area; it also schedules training for staff on new materials and safety handling.
- Paraloid B-72
- BEVA 371
- sturgeon glue
- funori
Operation
Real-time material viability checks
During treatment, AI interprets FTIR spectra, DSC data, viscosity curves, peel strength, substrate hydration, and micrographic observations; it assesses delamination risk, consolidant penetration depth, reversibility likelihood, and cumulative material aging before each procedural step, with alerts for substrate-specific interactions such as oak aging and linen embrittlement.
- Paraloid B-72
- BEVA 371
- funori
- Klucel G
Maintenance
Ongoing performance monitoring
AI continuously monitors shelf-life, batch variation, humidity exposure, substrate aging indicators, and historical performance; it flags out-of-spec readings, records consolidation outcomes, and schedules refresh or replacement of materials with compatibility notes and vendor updates; it logs re-treatments and documents stabilization outcomes for archival condition records.
- Paraloid B-72
- BEVA 371
- Klucel G
Safety
Reversibility and ethics auditing
AI verifies reversibility potential across solvents and consolidation layers, audits consented procedures, maintains PACR-linked documentation, COSHH safety compliance, and signs off treatment iterations with the accredited conservator and client approvals; it also tracks training completion, PPE usage, and containment containment efficacy.
- reversibility testing protocol
- Paraloid B-72
Part by part
- 1
Paraloid B-72
- Watch for
- Chemical compatibility with substrate aging and solvent sensitivity under ambient and elevated humidity.
- AI action
- Detects mismatch in substrate response and prompts alternative adhesive choice.
- 2
BEVA 371
- Watch for
- Pressure-induced stress in flexible substrates; yellowing risk under certain solvents.
- AI action
- Triggers mixed-matrix approach or reduced concentration recommendations.
- 3
sturgeon glue
- Watch for
- Proteinaceous substrate risk; microbial growth indicators during storage.
- AI action
- Suggests protective consolidation schedule and humidity controls.
- 4
funori
- Watch for
- Temporary consolidant performance under UV exposure; swelling behavior.
- AI action
- Recommends test patch and chelator-free tweaks if needed.
- 5
Klucel G
- Watch for
- Hydroxypropyl cellulose sensitivity to moisture; viscosity drift.
- AI action
- Prompts viscosity adjustment plan and compatible solvents.
- 6
reversibility testing protocol
- Watch for
- Emerging reversibility concerns; long-term tracking of solvent-resin interactions.
- AI action
- Initiates reversibility baseline and flags when re-treatment is required.
Worked example
AI-assisted adhesive and consolidant selection
Illustrative scenario with sample data.
Private studio with mixed-collection program seeks non-destructive consolidation plan.
Signal in
materialHistory=engraved panel; substrate=oak; priorTreatments=paraloid IMP; spectra=FTIR-ATR; condition=delamination@edge
System out
Treatment proposal generated: BEVA 371 consolidation with reversibility notes; proposed test patch schedule; documented rationale and expected outcomes.
Human sign-off: Head Conservator reviews AI proposal, approves initial patch tests, and signs the treatment proposal.
How to roll it out
Phase 1
Weeks 1 to 4: Baseline
Establish data pipelines, calibrate spectroscopic sensors, and generate initial treatment rules.
Phase 2
Weeks 5 to 10: Trialing
Run bench tests with sample substrates, compare performance, refine AI thresholds, and document decisions.
Phase 3
Weeks 11 onward: Scale
Apply approved materials to projects and integrate AI outputs into condition reports and curation records.
Where people stay in charge
- Decisions on material selection must be reviewed and signed by a licensed conservator or accredited professional.
- All AI-generated recommendations must be reconciled with reversibility requirements and PACR accreditation standards.
- Solvent and material handling decisions must comply with COSHH and local safety regulations; final authorization rests with the supervising conservator.
Questions owners and crews ask
What data does the AI need to suggest adhesives and consolidants?
The AI uses condition reports, prior treatments, substrate type, spectral data, and environmental readings to propose compatible materials and note reversibility considerations.
Who approves AI-generated treatment proposals?
A licensed conservator validates AI-driven proposals, signs treatment proposals, and ensures alignment with ethics and accreditation guidelines.
How is reversibility assessed in practice?
Reversibility testing protocol runs standardized tests to verify removability under controlled conditions and documents any irreversible interactions.
How does AI handle safety and COSHH concerns?
AI flags potential hazards, links to COSHH data, and ensures operators follow approved handling and containment procedures with sign-off by the safety lead.