Why it matters
Without AI, manual monitoring can miss rising water, delayed pump starts, or improper discharge handling, risking trench instability, service interference, and backlogs in post-excavation assessment and archive deposition.
From design to safety
Design
AI informs dewatering design parameters
AI reads historical water rise data, soil permeability, wellpoint layouts, and discharge constraints to shape the Written Scheme of Investigation and trench dewatering plan, incorporating groundwater model outputs, seasonal rainfall patterns, and TAN limits for discharge. It evaluates power supply capacity, generator siting, and backup pump options.
- Sump chamber
- Submersible pump
- Wellpoints
- Discharge consents
Build
Realtime dewatering setup and validation
AI integrates live water level sensors, pump runtimes, and discharge records to validate dewatering design during trench setup and initial excavation, while logging trench depth, shale horizons, and sediment strata that influence drawdown efficiency.
- Sump chamber
- Submersible pump
- Wellpoints
- Water levels
Operation
Continuous control and monitoring
AI continuously compares sensor readings, pump performance, rainfall forecasts, and site progress to optimize pumping cycles, manage discharge compliance, and generate daily welfare reports for client review.
- Sump chamber
- Submersible pump
- Wellpoints
- Water levels
Maintenance
Maintenance triggers and record-keeping
AI flags pump maintenance needs, sensor calibration drift, replacement part schedules, and discharge consent renewals, generating service orders, updating the post-excavation assessment records, and maintaining asset history.
- Submersible pump
- Discharge consents
- Water levels
- Silt
Safety
Buried services and safety compliance
AI cross-checks dewatering activity against buried services data, trench stability indicators, contamination risk, and UXO status, supporting planning compliance, site welfare, and safety audits with auditable trails.
- Discharge consents
- Silt
- Water levels
- Wellpoints
Part by part
- 1
Sump chamber
- Watch for
- Sump level rising beyond threshold; pump overheating or clogging.
- AI action
- Read sump level and pump vibration; trigger automatic pump restart or alarm to supervisor.
- 2
Submersible pump
- Watch for
- Vibration anomalies or reduced flow indicating impeller clog or motor fault.
- AI action
- Monitor flow rate and motor temperature; issue maintenance task and temporary decommission alert.
- 3
Wellpoints
- Watch for
- Insufficient drawdown or wellpoint clogging reducing dewatering efficiency.
- AI action
- Evaluate suction head data and slurry load; optimize point spacing and pump scheduling.
- 4
Discharge consents
- Watch for
- Consent expiry or exceedance of authorised discharge parameters.
- AI action
- Validate discharge data against consent limits; generate renewal/notification to compliance lead.
- 5
Silt
- Watch for
- Silt buildup in sumps affecting separation and clog risk.
- AI action
- Assess sediment load and inform timing of silt removal and filter changes.
- 6
Water levels
- Watch for
- Unusual water table fluctuations indicating drainage failure.
- AI action
- Flag anomalies; adjust pump sequence and notify supervisor and client reporting.
Worked example
Site dewatering for a 6.2 ha development—periods finds including Late Bronze Age layers
Illustrative scenario with sample data.
During trenching, sustained rainfall raises water levels. AI coordinates dewatering to maintain safe working depths while preserving stratigraphic integrity.
Signal in
sensor_water_level_mm: 210, pump_run_hours: 6, discharge_flow_L_s: 15, consent_status: active, silt_saturation: 0.65
System out
alert: rising water detected; action: initiate ramp-down of high-flow pumps; notify supervisor; record in context sheets and post-excavation plan.
Human sign-off: Project manager reviews alert, signs off on extended dewatering plan and updates HER with discharge and safety notes.
How to roll it out
Phase 1
Weeks 1 to 2
Baseline dewatering sensors installed; initial thresholds defined; data pipelines tested.
Phase 2
Weeks 3 to 6
Automated pump control and discharge logging active; contingency plans rehearsed.
Phase 3
Weeks 7 onward
Full integration with site reporting, safety audits, and post-excavation assessment workflow.
Where people stay in charge
- Decisions about discharge limits, pump selection, and dewatering thresholds must be made by a qualified person, based on the WSI and consent conditions because incorrect control risks safety and compliance.
- The system may propose changes, but only a licensed supervisor or planning archaeologist signs off on operational adjustments that affect site welfare and HER obligations.
- All data access, sensor deployments, and discharge actions require auditable records and sign-off to support post-excavation assessment and regulatory review.
Questions owners and crews ask
What triggers AI to adjust pump operation on site?
AI reads real-time water level, pump status, and discharge constraints to adjust pump speed, switch units, or pause pumping; a qualified supervisor approves any major change.
How are discharge consents tracked and updated?
AI monitors consent parameters, expiry dates, and actual discharge data, then generates renewal reminders and flags any exceedances to the compliance lead for sign-off.
Who reviews dewatering decisions before implementation?
The project manager or planning archaeologist reviews operational AI recommendations, ensuring alignment with the WSI, HER obligations, and site safety planning.
How is data archived after dewatering ends?
All sensor logs, pump maintenance, discharge records, and context sheets are archived in the project archive with references to the written scheme and post-excavation assessment.