STS / SMART TUNNEL SCANNER

Smart Tunnel Scanning Solution

Scan the full line during daily train operations

STS combines onboard laser scans with TIDS positioning to detect clearance intrusions, identify tunnel-wall defects and build a 3D record during service. It produces records in 5 m segments without a dedicated inspection vehicle.

02 / STS

From periodic surveys to routine scanning

Limited possession windows

Short service breaks favour sampled sections, leaving weeks between observations.

Subtle early defects

Small changes in intrusions, seepage, cracks and spalling can escape visual inspection.

Difficult follow-up

Without mileage, coordinates and raw point clouds, verification often requires another visit.

DimensionManual inspectionSTS
FrequencyWeekly to monthly section samplingPer-run scanning during daily operation
Defect locationManually estimated locationMileage and 3D coordinates; centimetre-scale mileage positioning requires project validation
Retained evidenceSeparate photos and notesContinuous point-cloud records for review and comparison
Track possessionRequires a service breakRuns with service trains without a separate survey possession
03 / STS

System components and scanning

Onboard equipment

Onboard equipment comprises the STS host, laser scanner, time-alignment unit and DPP status panel for processing and storage, profile capture, aligned records and health indication.

Onboard TIDS

Supplies position, speed and direction over Ethernet with link-health monitoring. Positioning and timing for trains without TIDS require separate confirmation.

STS cloud platform

Overview, defect statistics, trends, archives and downloads.

Scanning principle

Align 200 Hz tunnel profiles with TIDS position and speed by timestamp, split into 5 m segments and evaluate each to create a record with mileage and coordinates.

How coverage accumulates

Each scan captures one profile; train movement covers the tunnel length. Repeated runs increase coverage density, and lower speeds produce denser point clouds. This allows each run to scan the full line.

Vehicle integration

Choose roof or coupler mounting for the vehicle and clearance conditions. GB 50517 and EN 50155 inform the design; project documentation determines conformity.

04 / STS

Clearance intrusion detection

  1. Detection method

    Scan the track area in 5 m segments, compare the point cloud with clearance models and trigger an alert on intrusion.

  2. Three clearance models checked in parallel

    Structural clearance, kinematic envelope and a user-defined clearance envelope.

  3. Straight and curved track adaptation

    Identify track curvature and adjust scan parameters to maintain comparable judgement accuracy on straight and curved sections.

  4. Output

    Intrusion depth and 3D coordinates; normal regions are black and intrusions highlighted red.

  5. Alarm fields

    Timestamp, triggered envelope type, line section, precise mileage, direction, track Link ID, intrusion coordinates and raw PCD file.

Tunnel cross-section point-cloud example

The system records clearance decisions and evidence. The operator determines object nature and action under site procedures.

05 / STS

Tunnel-wall defect detection

DefectDetection scopeOutput
SeepageSeepage, wet patches and water streaksMileage and 3D coordinates for tracking changes in extent
CracksContinuous and network crackingDirection, length, width and density
Concrete spallingBroad spalling, fallen pieces and rockfall risk locationsDefect report and falling-piece risk indication

Enable the three tasks independently by maintenance priorities. Store each defect type in a separate directory; alerts retain time, type, station section, mileage, geographic coordinates and PCD evidence.

Width detection responds strongly to surface-wet cracks but has limited performance on dry cracks. Alerts locate cracks; specialist assessment is needed for depth and structural significance.

06 / STS

3D tunnel records

Processing workflow

Export raw points as standard PCD, align them with CSV mileage logs and build a full-line 3D record using straightened modelling. Each PCD holds accumulated points for one TIDS interval to preserve sequence.

Four uses

UseDescription
Deformation monitoringCompare the same section across periods to observe deformation trends
Lifecycle recordsTraceable records from handover through maintenance
Defect visualisationShow spatial distribution and development of cracking and leakage
Decision supportInform maintenance plans, spares planning and safety assessment
Illustrative 3D tunnel structure

Deformation comparison accuracy measures relative change at one location across periods. Minimum detectable target size, listed under performance, is a separate measure and must not be substituted for it.

07 / STS

Detection performance by speed

Minimum detectable intrusion, seepage and damage size

SpeedStraight trackCurve (R=300 m)
20 km/h≥83×47 mm≥84×47 mm
40 km/h≥167×47 mm≥168×47 mm
60 km/h≥250×47 mm≥252×47 mm
80 km/h≥333×47 mm≥336×47 mm

Minimum continuous-crack region

SpeedMinimum continuous-crack region
20 km/h≥165×20 mm
40 km/h≥335×20 mm
60 km/h≥500×20 mm
80 km/h≥667×20 mm

Targets must span ≥3 radial frames and ≥5 points per frame. Validate reflective and absorbent surfaces onsite. These test figures depend on speed, material, mounting angle and tunnel conditions.

08 / STS

Cloud review and traceability

Deployment

A dedicated cloud server is isolated from other product services, supports 24-hour access and allows expansion.

Three entry points

Overview, statistics and raw-data downloads.

Four data functions

FunctionDescription
Multidimensional searchFilter by type, time and location; inspect all alarm fields and download linked PCD
Historical retentionRetain approximately three years of intrusion, seepage, crack and spalling logs
Dual-window comparisonCompare defect distributions and counts across periods and sections side by side
Batch downloadDownload five raw PCD files per batch; buffer offline and resume transfers; export xlsx, csv and pcd

Statistical views

Monthly top-five sections by defect share, bar charts comparing counts by date and section, and a 30-day trend overlaying total events with intrusion growth.

Offline continuity

Historical results and logs remain available in the cloud while the train is offline. Missing results upload automatically when it reconnects.

Last updated