π Lesson 16
D5
Importing Drone DSMs into Deswik.CAD for Updated Bench Designs
Importing drone-derived digital surface models (DSMs) into Deswik.CAD lets engineers update mine bench designs using real, up-to-date 3D terrain data captured by drones.
π― Learning Objectives
- β Explain the geospatial preprocessing steps required before importing a drone DSM into Deswik.CAD
- β Apply coordinate system transformation and vertical datum correction to align drone DSMs with mine site control networks
- β Design an updated bench layout in Deswik.CAD using imported DSM contours and validate against design tolerances
- β Analyze volume discrepancies between original design surfaces and updated DSM-based surfaces to quantify overbreak or underbreak
- β Evaluate DSM resolution and accuracy against Deswik.CAD modeling requirements for bench design fidelity
π Why This Matters
In modern open-pit mines, bench geometry drifts due to weathering, blast movement, and operational wear β sometimes by >0.5 m per month. Relying on outdated CAD models leads to inaccurate drill pattern placement, suboptimal loading, and unsafe highwalls. Drone-based DSMs capture centimetre-accurate surface changes in hours, not weeks. Importing them into Deswik.CAD closes the loop between reality and design β enabling rapid, evidence-based re-benching, improved safety compliance, and precise production forecasting. This is no longer 'nice to have'; itβs how Tier-1 miners like Rio Tinto and BHP now manage their short-term planning cycles.
π Core Principles
Successful DSM import hinges on three interdependent layers: (1) Geospatial integrity β drone data must be captured in the same projected coordinate reference system (CRS) as the mineβs master control network (e.g., WGS84 UTM Zone 56S + AHD vertical datum), with ground control points (GCPs) achieving β€5 cm horizontal and β€3 cm vertical RMSE; (2) Data structure compatibility β Deswik.CAD v2023+ natively supports GeoTIFF DSMs and LAS/LAZ point clouds, but requires raster resampling to β€0.5 m grid cell size for bench-scale fidelity; (3) Engineering interpretation β raw DSMs include vegetation, equipment, and temporary stockpiles; users must apply morphological filtering or manual masking *before* import to isolate the true bench surface. Failure at any layer introduces cascading errors in slope angle calculation, berm width validation, and haul truck path clearance analysis.
π DSM Grid Resolution vs. Bench Design Fidelity
The maximum allowable DSM grid cell size (G) is governed by the smallest geometric feature needing representation β typically berm width (B) or catch bench depth (D). A conservative rule ensures G β€ B/4 to resolve berms without stair-stepping artifacts in contour generation.
Minimum DSM Grid Resolution Rule
G_min = B / 4Determines the coarsest acceptable grid cell size (in metres) to reliably represent a designed berm width (B) without geometric aliasing in contour generation.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| G_min | Minimum grid cell size | m | Largest permissible raster resolution for accurate bench feature representation |
| B | Designed berm width | m | Width of safety or catch berm as specified in engineering design |
Typical Ranges:
Primary haul road berms: 4.0 β 8.0 m
Catch berms on 45Β° slopes: 3.0 β 6.0 m
π‘ Worked Example
Problem: A mineβs primary catch berm is designed at 6.0 m wide. The drone DSM was collected at 1.2 m GSD (ground sample distance) and exported as a 1.5 m GeoTIFF grid. Does this meet Deswik.CAD bench design fidelity requirements?
1.
Step 1: Calculate minimum required grid resolution: G_min = B / 4 = 6.0 m / 4 = 1.5 m
2.
Step 2: Compare with actual grid: G_actual = 1.5 m β meets threshold exactly
3.
Step 3: Verify vertical accuracy: For 1.5 m grid, vertical RMSE must be β€0.15 m (per Deswik Best Practice Guide v4.2, Sec 7.3) β drone survey achieved 0.11 m RMSE via 12 GCPs
Answer:
The 1.5 m DSM grid meets the minimum fidelity requirement for a 6.0 m berm and satisfies vertical accuracy standards. However, for future highwall stability analysis, a 0.75 m grid is recommended.
ποΈ Real-World Application
At Newmontβs Boddington Mine (WA), monthly drone DSMs (collected at 0.08 m GSD using WingtraOne VTOL with PPK GNSS) were imported into Deswik.CAD to update the Stage 7 East pit benches. Pre-import, GCPs were surveyed to AHD93 vertical datum using Trimble R12 receivers (2 cm vertical RMSE). DSMs were clipped to pit limits, filtered to remove haul trucks using LAStools' lasheight, and resampled to 0.5 m grid. Within Deswik.CAD, the DSM was converted to a TIN surface, then used to regenerate 5 m contour lines. Engineers identified a 0.8 m overbreak along 120 m of the SW highwall β triggering a revised scaling design and preventing a potential geotechnical hazard. Volume reconciliation showed 23,500 mΒ³ of unplanned waste β directly fed into the next monthβs production schedule.
π§ Interactive Calculator
π§ Open Mine Drone-Based Surveying & Inspection Calculatorπ Case Connection
π Coal Mine Haul Road Surface Degradation Analysis
Unplanned truck tire failures due to undetected potholes and rutting; manual road surveys occurred only quarterly