Why it matters
Without AI support, dye migration, uneven cleaning, solvent mis-use, and unpredictable drying can harm textiles, causing irreversible colour loss and ethical concerns over material integrity.
From design to safety
Design
AI-guided treatment planning
The system reviews textile type, fibre content, dye media, historical surface finishes, and known degradation modes to propose safe wet cleaning methods, compatible surfactants, and a reversible, documented plan linked to the object record.
- wet cleaning bath system
- dye bleed test kit
- surfactant concentration sensor
Build
Bench setup and validation
Specification approves dedicated benches, PPE, DI water loop, dye bleed fixtures, and sensor calibration routines to validate cleaning parameters against historical references and PACR guidelines.
- deionised water purifier
- surfactant concentration sensor
- colour fastness specimen rack
Operation
Real-time risk monitoring
During cleaning, AI tracks bath parameters, dye bleed test results, surfactant usage, and drying humidity to trigger aborts, protocol updates, or documentation entries in the treatment log.
- wet cleaning bath system
- deionised water purifier
- colour fastness specimen rack
Maintenance
Calibration and records
Regular calibration of baths and sensors, stock checks for surfactants, replacement DI water cartridges, and maintenance of the DI loop with updated condition reports and lot tracing.
- dye bleed test kit
- surfactant concentration sensor
- colour fastness specimen rack
Safety
Ethics and compliance checks
AI logs all decisions, flags non-conformities to COSHH and PACR standards, and generates treatment reports, safety summaries, and archival-ready records for accreditation and insurance review.
- wet cleaning bath system
- deionised water purifier
- colour fastness specimen rack
Part by part
- 1
wet cleaning bath system
- Watch for
- Bleed patterns and over-cleaning signs on test swatches; dye migration toward adjacent areas; loss of texture under bath.
- AI action
- Detect dye migration thresholds and halt bath when approaching limits; trigger automated safety log entry.
- 2
dye bleed test kit
- Watch for
- Unstable dye transfer indicators or inconsistent colour release across swatches; sudden spikes in bleed index.
- AI action
- Flag anomalies and prompt retests before proceeding; log results in treatment plan.
- 3
surfactant concentration sensor
- Watch for
- Concentration drift or micelle formation beyond safe range; incomplete rinse indicators.
- AI action
- Adjust bath formulation and log parameter changes; alert operator if thresholds exceeded.
- 4
deionised water purifier
- Watch for
- Water conductivity or total dissolved solids rising above limits; cartridge degradation.
- AI action
- Maintain DI water purity, trigger maintenance if out of spec, record filter changes.
- 5
drying chamber with humidity control
- Watch for
- Elevated residual moisture or uneven air flow; slow steady-drying deviations.
- AI action
- Modulate humidity and airflow to meet target drying curve; document stepwise progress.
- 6
colour fastness specimen rack
- Watch for
- Inconsistent colourfastness readings across samples; mislabelled specimens.
- AI action
- Flag to conservator and recommend re-testing protocol; attach results to the object record.
Worked example
Mitigating dye bleed risk before full-scale wet cleaning
Illustrative scenario with sample data.
A private studio assesses a mid-19th-century silk velvet with unknown dye colourfastness.
Signal in
fibreType= silk; dyeBinder= mordant; bathTemp= 22C; surfactant= nonionic; w_pH= 6.8; sampleSwatches= 5; readingData= {bleedIndex:0.12, conductivity: 2.1 μS/cm, moisture: 12%}
System out
AI recommends staged cleaning with low surfactant concentration, extended dye bleed testing, and controlled drying parameters; a formal treatment proposal and risk log are generated.
Human sign-off: Head conservator approves the proposed staged protocol and signs the treatment proposal.
How to roll it out
Phase 1
Weeks 1 to 2
Baseline setup of DI loop, bath calibration, dye bleed testing workflow, and initial risk scoring templates.
Phase 2
Weeks 3 to 6
Pilot on a small textile sample set; refine AI thresholds, decision points, and reporting templates for client documentation.
Phase 3
Weeks 7 to 12
Full integration into standard operating procedures, archival filing, and examiner-reviewed client-facing reports.
Where people stay in charge
- All treatment decisions remain with the accredited conservator or designated supervisor familiar with PACR standards.
- AI can propose adjustments but cannot authorize dye-related changes without human sign-off.
- No irreversible changes should be performed without explicit ethical and archival approval.
Questions owners and crews ask
What data does the AI use to assess dye bleed risk in textiles?
The system analyzes fibre type, dye media, bath composition, temperature, pH, sample swatch bleed readings, and historical colourfastness records to gauge risk and guide steps.
How does the AI handle COSHH and solvent safety in wet cleaning?
It monitors chemical usage, exposure thresholds, ventilation status, and non-compliant formulations, routing decisions to the conservator for final approval and logging in the safety dossier.
What records are created for treatment proposals and condition reports?
AI generates a treatment proposal, stepwise protocol, risk assessment, and post-treatment condition report linked to the object record with audit trails.
How is client privacy and archival data protected?
Access controls, encrypted logs, and multifactor authentication ensure sensitive conservation data and client information remain secure and traceable.