Why it matters
Without AI, misconfigured sieving stations risk lost small finds, delayed processing, and site backlogs due to manual checks, inconsistent residue handling, and non-optimized recovery rates needing rework.
From design to safety
Design
AI-enabled station design planning
The AI ingests the written scheme of investigation, planned trench layouts, expected artifact sizes, and water considerations to propose sieve configurations and mesh ladders that align with context recording requirements and archive deposition workflows.
- sieves
- mesh sizes
- water supply
- sieving rotas
Build
Site setup and equipment alignment
The system validates water access, sieve frames, and rotas against trench dimensions, ensuring correct placement, feed rates, and filtration sequencing prior to first sieving pass, with field team checks logged.
- water supply
- sieving rotas
- sieves
Operation
Real-time sieving optimization
It reads live throughput, residuals, and recovery rates, adjusting mesh selection and water pressure to maximise efficient recovery while preserving context sheets and provenance notes linked to each trench.
- mesh sizes
- residues
- recovery rates
Maintenance
Calibration and throughput auditing
The AI schedules regular calibration of mesh counts, reviews residue handling, and logs recovery performance for post-excavation assessment, including chain-of-custody notes and archive deposition readiness.
- sieves
- recovery rates
- residues
Safety
Buried services and trench risk checks
It cross-references trench maps with rotor rotations and water delivery lines to prevent accidental service strikes, assess trench stability, and ensure welfare access remains clear during operations.
- water supply
- sieving rotas
- sieves
Part by part
- 1
sieves
- Watch for
- Wear or clogging affecting filtration efficiency and context capture.
- AI action
- Detect clogging trends and trigger rotor schedule adjustment.
- 2
mesh sizes
- Watch for
- Mesh wear or misalignment reducing recovery of small finds.
- AI action
- Flag degraded mesh and propose size switch with readjusted flow.
- 3
water supply
- Watch for
- Insufficient flow or contamination risk in sieving water.
- AI action
- Alert to pressure drop and switch to alternate source or treatment.
- 4
sieving rotas
- Watch for
- Incorrect rosters leading to uneven processing across trenches.
- AI action
- Rebalance rotas based on trench demand and throughput data.
- 5
residues
- Watch for
- Inadequate residue handling causing loss of material or cross-contamination.
- AI action
- Trigger checklist for containment, drying, and segregation.
- 6
recovery rates
- Watch for
- Unexpected drop in recovered material versus expectation.
- AI action
- Adjust sieve size and water throughput to restore performance.
Worked example
Catching artifacts early in a 14-ha development
Illustrative scenario with sample data.
During a 14-hectare, multi-period site, AI-assisted wet sieving coordinates water, mesh selection, and rotas to preserve small finds during trench strip-map and sample phases.
Signal in
siteAreaHa:14; periodsFound: medieval, roman; waterPressureBar:2.5; meshSizes: [2mm, 4mm, 8mm]; rotorSchedule: morning, afternoon; residuesCollected: light, medium; throughputsPerHour: [120, 95, 110]; trenchCount: 18; contextSheetsLinked: true
System out
AI-recommended mesh progression: start with 2mm for fine filtration, switch to 4mm in mid-workshift; rotor adjustments deployed; expect stabilized recovery, and log alert for deviation; archive deposition linked.
Human sign-off: Project manager reviews suggested configuration and approves deployment plan with the client’s heritage consultant; curator signs off after cross-checking context sheet references.
How to roll it out
Phase 1
Weeks 1 to 2: Baseline
Establish baseline sieve setup, mesh ladder, water supply, and rotor wavelengths across one trench for validation with computable tolerance bands.
Phase 2
Weeks 3 to 6: Expansion
Scale AI controls to additional trenches, validate multiple periods, and integrate with context sheets and archive deposition workflow with duplicate checks.
Phase 3
Weeks 7 to 12: Optimisation
Fine-tune throughputs, residues handling, and recovery data, aligning with post-excavation assessment requirements and HER reporting templates.
Where people stay in charge
- Decisions on mesh sizes, water pressures, and rotor schedules must be approved by the site supervisor or project manager with sign-off from the heritage consultant.
- All changes affecting context recording, watch status, or recovery rates require documented justification and HER update before implementation.
- Changes to waste handling, residue segregation, or archive deposition sequencing must be reviewed by the curator and logged in the HER.
Questions owners and crews ask
What data should we feed the AI wet sieving station planning system?
Feed trench dimensions, mesh sizes, water supply pressure, rotor rosters, throughput, residue categories, and context sheet references to enable accurate optimization and traceable decisions.
How does AI handle safety during sieving operations?
AI monitors water and rotor pressures, trench access, and service maps to alert teams of unsafe conditions and trigger protective measures.
Who signs off changes to the sieving protocol?
The site supervisor signs off on operational changes; the project manager confirms with the heritage consultant and curator if required.
What records are produced by the AI station for post-excavation analysis?
Automated sieving logs, mesh usage histories, water quality readings, residue handling notes, and recovery rate reports linked to context sheets and archive deposition.