Why it matters
Without this system, unexpected groundwater shifts, misinterpreted timber remains, or unrecorded finds can stall build, risk trench stability, or complicate post-excavation archiving and reporting.
From design to safety
Design
AI assists design risk assessment
Reads terrain models, historical records, WSI, trench design constraints, utility plans, stratigraphic indicators from borehole logs, and Lidar-derived microrelief; quantifies risk and proposes trench widths, shoring needs, sampling lanes, access points, and ramped entry points.
- waterlogged deposits
- timber structures
- deep sequences
Build
AI guides trench layout and safety
Interfaces with site layout data, topographic surveys, and utility maps to fix trench widths, method statements, shoring configurations, dewatering sump placement, and on-site sampling sequences; integrates phased daylighting and spoil management.
- waterlogged deposits
- shoring
- deep sequences
Operation
AI coordinates recording and processing
Monitors context sheets, Harris matrices, strip maps, and sample logs to plan on-site processing, lab intake, and client reporting with traceable workflows; schedules finds processing and on-site conservation checks.
- sampling
- conservation
- timber structures
Maintenance
AI orchestrates calibration and archiving
Reads equipment calibration logs, moisture readings, conservation notes, and post-excavation assessments to trigger calibration, conservation adjustments, and archive deposition scheduling; tracks lab turnaround and digital archive readiness.
- conservation
- sampling
- waterlogged deposits
Safety
AI sustains trench stability and safety
Watches trench sensors, buried services, contamination, and UXO alerts; triggers site-wide pauses, service alerts, engineer notification, welfare checks, and PPE compliance reporting.
- shoring
- waterlogged deposits
- deep sequences
Part by part
- 1
waterlogged deposits
- Watch for
- subsurface moisture surge, rising groundwater, anaerobic decay patterns, sudden seepage, and oxygen deficit signs in stratigraphic layers
- AI action
- reads groundwater sensors and soil moisture logs; triggers halting, backfilling, or drainage actions
- 2
timber structures
- Watch for
- timber collapse indicators, deformation of timbers, rot fronts, or hidden load-bearing joints
- AI action
- analyzes timbers in situ via 3D scanning; recommends stabilization or protective isolation measures
- 3
shoring
- Watch for
- shoring integrity loss, panel displacement, trench width creep, scaffold misalignment
- AI action
- monitors shoring strain gauges and spacing; initiates reinforcement or corrective actions
- 4
deep sequences
- Watch for
- unseen stratigraphic boundaries, sequence gaps, late-context deposition risk
- AI action
- reads stratigraphic logs and depth sensors; adjusts excavation plan and sampling lanes
- 5
sampling
- Watch for
- sample contamination risk, lost context, mislabelled jars or bags
- AI action
- triggers sample chain-of-custody checks and aligns lab intake with context sheets
- 6
conservation
- Watch for
- deterioration of fragile finds, need for chemical stabilization, storage label mismatches
- AI action
- issues conservation task orders and updates archive metadata
Worked example
Catching a buried timber raft before water impermeability halts trenching
Illustrative scenario with sample data.
A 0.9 ha urban site with waterlogged deposits and deep timber features requires careful planning to preserve timber structures while meeting WSI and HER constraints.
Signal in
water table reading=1.2 m, soil moisture=28%, depth of logs=0.9 m, timber age estimate=600-800 years, context sheets pending, shale layer at 1.6 m; additional borehole log data includes 4 logs with resistivity anomalies and 2 GPS-controlled permits.
System out
AI flags high groundwater risk, recommends enhanced shoring and temporary dewatering; updates trench plan and sampling lanes; alerts conservation team for timber preservation; generates a revised WSI map and dewatering protocol.
Human sign-off: Project manager approves revised trench design and dewatering plan; heritage consultant signs off on timber preservation approach; site supervisor implements changes; HER officer approves the updated WSI and sampling plan.
How to roll it out
Phase 1
Weeks 1 to 2
Baseline data models, site setup plans, and sampling strategy aligned with WSI and utility maps.
Phase 2
Weeks 3 to 6
Live trench monitoring integrates with ESRI-based layout, with dynamic shoring adjustments and interim reporting.
Phase 3
Weeks 7 onward
Post-excavation workflow enabled: post-excavation assessment, archive deposition readiness, and conservation status tracking.
Where people stay in charge
- Decisions affecting structural safety, dewatering, or deep sequence interpretation must be signed off by a qualified site engineer or project heritage lead.
- Legal and regulatory compliance decisions require the planning archaeologist or local authority HER officer to authorize.
- Any changes to the Written Scheme of Investigation must be reviewed and approved by the client’s heritage consultant.
Questions owners and crews ask
What kinds of data should we feed the AI for urban waterlogged sites?
Feed topographic data, utility maps, groundwater readings, soil moisture logs, WSI constraints, context sheets, and sampling plans; the AI relies on current, linked datasets for decisions.
How does the AI handle uncertainty in timber structures?
It flags uncertain timbers, proposes stabilization measures, and schedules specialist timber conservation reviews, ensuring decisions are validated by a qualified archaeologist and conservator.
Can the system trigger safety pauses automatically?
Yes, when trench sensors or detection of buried services indicate risk, it can pause operations and alert the supervisor for immediate action.
What records are generated for post-excavation archiving?
It creates context sheets, Harris matrices, plan records, sampling logs, conservation notes, and archive deposition metadata ready for HER and archive submission.