Why it matters
Without AI, buried services are missed or misread, causing service strikes, trench instability, delays in permit approvals, and post-excavation backlogs that ripple through schedules and budgets.
From design to safety
Design
AI guides risk-aware design decisions
The system reads utility drawings, all geophysical scans, and WSI inputs to compare planned trenches with known services, flag inconsistencies, prompt design-review updates, and log decisions in the Harris matrix for traceability.
- utility drawings
- signal generators
- permit-to-dig
Build
AI sequences trench setup checks
During site set-up, the AI cross-references cable avoidance tool data with live signals, validates trench methods, flags conflicting readings, and generates collaborative action orders to the machinist and site foreman.
- cable avoidance tools
- signal generators
- permit-to-dig
Operation
AI supports ongoing monitoring and updates
In excavation, the AI tracks sensor feeds, updates Harris matrix context, logs service proximity events, produces daily client briefings, and maintains permit status with an audit trail for regulatory review.
- cable avoidance tools
- signal generators
- hand digging
Maintenance
AI maintains data integrity and records
Post-excavation, the AI compiles service records, cross-checks with archive drawings and HER, flags discrepancies needing post-excavation assessment input, and stores copies in the project archive with version control.
- utility drawings
- permit-to-dig
- service strikes
Safety
AI enforces trench safety and compliance
Real-time proximity alerts and buried service rechecks trigger formal safety notices, contamination watches, and require qualified sign-off before any further work, with automatic resubmission to planning archaeologists.
- service strikes
- permit-to-dig
- hand digging
Part by part
- 1
cable avoidance tools
- Watch for
- sensor drift or misreadings that hide a live service near the trench edge
- AI action
- analyzes tool output and triggers immediate avoidance actions and a stop-work signal
- 2
signal generators
- Watch for
- ambiguous signal patterns indicating crossing cables or degraded coupling
- AI action
- aggregates multiple signals to confirm service path and update dig plan
- 3
utility drawings
- Watch for
- outdated or conflicting drawings with site reality
- AI action
- flags discrepancies and prompts revision documents for the WSI and HER records
- 4
permit-to-dig
- Watch for
- missing deconfliction steps or unclear permit conditions
- AI action
- verifies permit clauses against current trench plan and generates corrective action orders
- 5
hand digging
- Watch for
- unexpected features or deviations from planned context
- AI action
- triggers pause and requests supervisor review with updated context sheets
- 6
service strikes
- Watch for
- new contact with a suspected live service during excavation
- AI action
- issues containment protocol and updates risk register with service status
Worked example
Detecting a buried service during a camera survey on a 2.3ha site
Illustrative scenario with sample data.
A motorway-adjacent site with multiple trenches requires permit-to-dig under HER conditions during early WSI validation.
Signal in
site_id: A23-Delta, area_ha: 2.3, periods_found: roman?0; earthworks_start: 2026-11; utility_drawings_ver: 2026-06; readings: toolA: normal, toolB: drift, signal_gen: 0.8
System out
alert: possible services near trench 4, recommendation: pause, re-check with shielded probes, log to permit-to-dig and WSI
Human sign-off: Project manager approves pause and orders a re-survey before continuing; planning archaeologist signs off clarifications.
How to roll it out
Phase 1
Weeks 1 to 4
Baseline data and WSI checks align with existing utility drawings and initial permit conditions.
Phase 2
Weeks 5 to 10
Integrated monitoring expands to live trench work, with real-time signals feeding the permit-to-dig workflow.
Phase 3
Weeks 11 to 20
Full operation phase with automated reporting, audit trails, and archive deposition readiness.
Where people stay in charge
- Decisions affecting permit-to-dig obligations must be approved by the project manager and reviewed by the client’s heritage consultant.
- All safety actions require sign-off by a qualified person: the site supervisor or planning archaeologist when regulatory scope applies.
- Discrepancies between drawings and field reality require escalation to the HER officer and post-excavation advisor before proceeding.
Questions owners and crews ask
How does AI buried services detection reduce the risk of service strikes during trenching?
AI reads live sensor data, cross-checks with utility drawings, and issues alerts or stop-work orders when a service path is uncertain, ensuring supervisor review and documented evidence before continuing.
What records are generated for permit-to-dig control in archaeology projects?
The system creates updated dig plans, service proximity logs, revised WSI inputs, permit-to-dig action logs, and a traceable permit version in the archive for client and HER review.
How are discrepancies between drawings and field reality handled?
Discrepancies trigger an escalation workflow to update WSI, issue revised context sheets, obtain sign-off from the planning archaeologist, and log changes in the Harris matrix.
Who approves AI-generated safety actions in trench work?
Qualified persons sign off, typically the site supervisor or planning archaeologist, with final approval from the client’s heritage consultant where required, and regulatory sign-off where applicable.