Drone Inspection Specialist
Expert in drone-based infrastructure inspection with computer vision, thermal analysis, and 3D reconstruction for insurance, property assessment, and environmental monitoring.
Decision Tree: When to Use This Skill
User mentions drones/UAV?
├─ YES → Is it about inspection or assessment of something?
│ ├─ Fire detection, smoke, thermal hotspots → THIS SKILL
│ ├─ Roof damage, hail, shingles → THIS SKILL
│ ├─ Property/insurance assessment → THIS SKILL
│ ├─ 3D reconstruction for measurement → THIS SKILL
│ ├─ Wildfire risk, defensible space → THIS SKILL
│ └─ NO (flight control, navigation, general CV) → drone-cv-expert
└─ NO → Is it about fire/roof/property assessment without drones?
├─ YES → Still use THIS SKILL (methods apply)
└─ NO → Different skill needed
Core Competencies
Fire Detection & Wildfire Risk
- Multi-Modal Detection: RGB smoke + thermal hotspot fusion
- Precondition Assessment: NDVI, fuel load, vegetation density
- Defensible Space: CAL FIRE/NFPA 1144 compliance evaluation
- Progression Tracking: Spread rate, direction prediction
Roof & Structural Inspection
- Damage Detection: Cracks, missing shingles, wear, ponding
- Hail Analysis: Impact pattern recognition, size estimation
- Thermal Analysis: Moisture detection, insulation gaps, HVAC leaks
- Material Classification: Asphalt, metal, tile, slate identification
3D Reconstruction (Gaussian Splatting)
- Pipeline: Video → COLMAP SfM → 3DGS training → Web viewer
- Measurements: Roof area, damage dimensions, property bounds
- Change Detection: Before/after comparison for claims
Insurance & Reinsurance
- Claim Packaging: Documentation meeting industry standards
- Risk Modeling: Catastrophe models, loss distributions
- Precondition Data: Satellite + drone + ground integration
Anti-Patterns to Avoid
1. "Single-Sensor Dependence"
Wrong: Using only RGB for fire detection.
Right: Multi-modal fusion (RGB + thermal) for high-confidence alerts.
| Detection Source |
Confidence |
Action |
| Thermal fire only |
70% |
Alert + verify |
| RGB smoke only |
60% |
Alert + investigate |
| Thermal + RGB |
95% |
Confirmed fire |
2. "Ignoring Hail Pattern"
Wrong: Counting damage without analyzing spatial distribution.
Right: True hail damage has RANDOM distribution. Linear or clustered patterns indicate other causes (foot traffic, age).
3. "Thermal Temperature Trust"
Wrong: Using raw thermal values without calibration.
Right: Account for:
- Emissivity of materials (roof = 0.9-0.95)
- Atmospheric transmission (humidity, distance)
- Reflected temperature from surroundings
- Time of day (thermal lag)
4. "3DGS Frame Overload"
Wrong: Extracting every frame from drone video.
Right: Extract 2-3 fps with 80% overlap. More frames ≠ better reconstruction.
| Video FPS |
Extract Rate |
Result |
| 30 |
30 (all) |
Redundant, slow processing |
| 30 |
2-3 |
Optimal quality/speed |
| 30 |
0.5 |
Insufficient overlap |
5. "Insurance Claim Speculation"
Wrong: Estimating costs without material identification.
Right: Identify material → Apply correct cost matrix.
| Material |
Repair $/sqft |
Replace $/sqft |
| Asphalt shingle |
$5-10 |
$3-7 |
| Metal |
$10-15 |
$8-14 |
| Tile |
$12-20 |
$10-18 |
| Slate |
$20-40 |
$15-30 |
6. "Defensible Space Zone Confusion"
Wrong: Treating all vegetation equally regardless of distance.
Right: CAL FIRE zones have different requirements:
| Zone |
Distance |
Requirement |
| 0 |
0-5 ft |
Ember-resistant (no combustibles) |
| 1 |
5-30 ft |
Lean, clean, green (spaced trees) |
| 2 |
30-100 ft |
Reduced fuel (selective thinning) |
Data Collection Strategy
Satellite Data (Regional Context)
- Sentinel-2: 10m resolution, NDVI, fuel moisture (SWIR bands)
- Landsat-8: 30m resolution, historical baseline, thermal band
- Planet: 3m resolution daily, change detection
- Application: Regional risk mapping, before/after events
Drone Data (Property Detail)
- RGB Mapping: 2-5cm GSD, orthomosaic, 3D model
- Thermal Survey: Moisture detection, heat signatures
- Close Inspection: Damage documentation, detail photos
- Application: Individual property assessment
Ground Truth
- Slope Measurement: GPS transects for topographic risk
- Soil Sampling: Moisture content for fire risk
- Material Verification: Confirm roof type
- Application: Calibration and validation
Quick Reference Tables
Fire Detection Confidence Levels
| Signal Combination |
Confidence |
Alert Priority |
| Thermal >150°C + Smoke |
95% |
CRITICAL |
| Thermal fire model |
80% |
HIGH |
| Hotspot >80°C |
70% |
MEDIUM |
| Smoke only |
60% |
MEDIUM |
| Hotspot 60-80°C |
50% |
LOW |
Roof Damage Severity
| Type |
Low |
Medium |
High |
Critical |
| Missing shingle |
- |
- |
Always |
- |
| Crack |
<1" |
1-3" |
>3" |
Multiple |
| Granule loss |
<10% |
10-30% |
>30% |
- |
| Ponding |
- |
Small |
Large |
Active leak |
Wildfire Risk Factors (Weighted)
| Factor |
Weight |
High Risk Indicators |
| Defensible space |
20% |
Non-compliant zones |
| Vegetation density |
20% |
NDVI >0.6, high fuel load |
| Slope |
15% |
>30% grade |
| Roof material |
10% |
Wood shake, Class C |
| Structure spacing |
10% |
<30ft between buildings |
| Access/egress |
10% |
Single road, narrow |
3DGS Quality Settings
| Quality Level |
Iterations |
Time |
Use Case |
| Preview |
7K |
5 min |
Quick check |
| Standard |
30K |
30 min |
General use |
| High |
50K |
60 min |
Documentation |
| Inspection |
100K |
3 hrs |
Damage measurement |
Reference Files
Detailed implementations in references/:
fire-detection.md - Multi-modal fire detection, thermal cameras, progression trackingroof-inspection.md - Damage detection, thermal analysis, material classificationinsurance-risk-assessment.md - Hail damage, wildfire risk, catastrophe modeling, reinsurancegaussian-splatting-3d.md - COLMAP pipeline, 3DGS training, inspection measurements
Integration Points
- drone-cv-expert: Flight control, navigation, general CV algorithms
- metal-shader-expert: GPU-accelerated 3DGS rendering
- collage-layout-expert: Visual report composition
- clip-aware-embeddings: Material/damage classification assistance
Insurance Workflow
1. Pre-Event Assessment (Underwriting)
├─ Satellite: Regional risk context
├─ Drone: Property-level risk factors
└─ Output: Risk score, premium factors
2. Post-Event Inspection (Claims)
├─ Drone survey: Damage documentation
├─ 3DGS: Measurements, change detection
└─ Output: Claim package, cost estimate
3. Portfolio Risk (Reinsurance)
├─ Aggregate: TIV, loss curves
├─ Model: AAL, PML, concentration
└─ Output: Treaty pricing, structure
Key Principle: Inspection accuracy depends on multi-source data fusion. Single-sensor assessments miss critical context. Always correlate drone findings with satellite baseline and weather data for defensible conclusions.
1---2name: drone-inspection-specialist3description: Advanced CV for infrastructure inspection including forest fire detection, wildfire precondition assessment, roof inspection, hail damage analysis, thermal imaging, and 3D Gaussian Splatting reconstruction. Expert in multi-modal detection, insurance risk modeling, and reinsurance data pipelines. Activate on "fire detection", "wildfire risk", "roof inspection", "hail damage", "thermal analysis", "Gaussian Splatting", "3DGS", "insurance inspection", "defensible space", "property assessment", "catastrophe modeling", "NDVI", "fuel load". NOT for general drone flight control, SLAM, path planning, or sensor fusion (use drone-cv-expert), GPU shader development (use metal-shader-expert), or generic object detection without inspection context (use clip-aware-embeddings).4license: Apache-2.05---6
7# Drone Inspection Specialist
8
9Expert in drone-based infrastructure inspection with computer vision, thermal analysis, and 3D reconstruction for insurance, property assessment, and environmental monitoring.
10
11## Decision Tree: When to Use This Skill
12
13```
14User mentions drones/UAV?
15├─ YES → Is it about inspection or assessment of something?
16│ ├─ Fire detection, smoke, thermal hotspots → THIS SKILL
17│ ├─ Roof damage, hail, shingles → THIS SKILL
18│ ├─ Property/insurance assessment → THIS SKILL
19│ ├─ 3D reconstruction for measurement → THIS SKILL
20│ ├─ Wildfire risk, defensible space → THIS SKILL
21│ └─ NO (flight control, navigation, general CV) → drone-cv-expert
22└─ NO → Is it about fire/roof/property assessment without drones?
23 ├─ YES → Still use THIS SKILL (methods apply)
24 └─ NO → Different skill needed
25```
26
27## Core Competencies
28
29### Fire Detection & Wildfire Risk
30- **Multi-Modal Detection**: RGB smoke + thermal hotspot fusion
31- **Precondition Assessment**: NDVI, fuel load, vegetation density
32- **Defensible Space**: CAL FIRE/NFPA 1144 compliance evaluation
33- **Progression Tracking**: Spread rate, direction prediction
34
35### Roof & Structural Inspection
36- **Damage Detection**: Cracks, missing shingles, wear, ponding
37- **Hail Analysis**: Impact pattern recognition, size estimation
38- **Thermal Analysis**: Moisture detection, insulation gaps, HVAC leaks
39- **Material Classification**: Asphalt, metal, tile, slate identification
40
41### 3D Reconstruction (Gaussian Splatting)
42- **Pipeline**: Video → COLMAP SfM → 3DGS training → Web viewer
43- **Measurements**: Roof area, damage dimensions, property bounds
44- **Change Detection**: Before/after comparison for claims
45
46### Insurance & Reinsurance
47- **Claim Packaging**: Documentation meeting industry standards
48- **Risk Modeling**: Catastrophe models, loss distributions
49- **Precondition Data**: Satellite + drone + ground integration
50
51## Anti-Patterns to Avoid
52
53### 1. "Single-Sensor Dependence"
54**Wrong**: Using only RGB for fire detection.
55**Right**: Multi-modal fusion (RGB + thermal) for high-confidence alerts.
56| Detection Source | Confidence | Action |
57|------------------|------------|--------|
58| Thermal fire only | 70% | Alert + verify |
59| RGB smoke only | 60% | Alert + investigate |
60| Thermal + RGB | 95% | Confirmed fire |
61
62### 2. "Ignoring Hail Pattern"
63**Wrong**: Counting damage without analyzing spatial distribution.
64**Right**: True hail damage has RANDOM distribution. Linear or clustered patterns indicate other causes (foot traffic, age).
65
66### 3. "Thermal Temperature Trust"
67**Wrong**: Using raw thermal values without calibration.
68**Right**: Account for:
69- Emissivity of materials (roof = 0.9-0.95)
70- Atmospheric transmission (humidity, distance)
71- Reflected temperature from surroundings
72- Time of day (thermal lag)
73
74### 4. "3DGS Frame Overload"
75**Wrong**: Extracting every frame from drone video.
76**Right**: Extract 2-3 fps with 80% overlap. More frames ≠ better reconstruction.
77| Video FPS | Extract Rate | Result |
78|-----------|--------------|--------|
79| 30 | 30 (all) | Redundant, slow processing |
80| 30 | 2-3 | Optimal quality/speed |
81| 30 | 0.5 | Insufficient overlap |
82
83### 5. "Insurance Claim Speculation"
84**Wrong**: Estimating costs without material identification.
85**Right**: Identify material → Apply correct cost matrix.
86| Material | Repair $/sqft | Replace $/sqft |
87|----------|--------------|----------------|
88| Asphalt shingle | $5-10 | $3-7 |
89| Metal | $10-15 | $8-14 |
90| Tile | $12-20 | $10-18 |
91| Slate | $20-40 | $15-30 |
92
93### 6. "Defensible Space Zone Confusion"
94**Wrong**: Treating all vegetation equally regardless of distance.
95**Right**: CAL FIRE zones have different requirements:
96| Zone | Distance | Requirement |
97|------|----------|-------------|
98| 0 | 0-5 ft | Ember-resistant (no combustibles) |
99| 1 | 5-30 ft | Lean, clean, green (spaced trees) |
100| 2 | 30-100 ft | Reduced fuel (selective thinning) |
101
102## Data Collection Strategy
103
104### Satellite Data (Regional Context)
105- **Sentinel-2**: 10m resolution, NDVI, fuel moisture (SWIR bands)
106- **Landsat-8**: 30m resolution, historical baseline, thermal band
107- **Planet**: 3m resolution daily, change detection
108- **Application**: Regional risk mapping, before/after events
109
110### Drone Data (Property Detail)
111- **RGB Mapping**: 2-5cm GSD, orthomosaic, 3D model
112- **Thermal Survey**: Moisture detection, heat signatures
113- **Close Inspection**: Damage documentation, detail photos
114- **Application**: Individual property assessment
115
116### Ground Truth
117- **Slope Measurement**: GPS transects for topographic risk
118- **Soil Sampling**: Moisture content for fire risk
119- **Material Verification**: Confirm roof type
120- **Application**: Calibration and validation
121
122## Quick Reference Tables
123
124### Fire Detection Confidence Levels
125| Signal Combination | Confidence | Alert Priority |
126|-------------------|------------|----------------|
127| Thermal >150°C + Smoke | 95% | CRITICAL |
128| Thermal fire model | 80% | HIGH |
129| Hotspot >80°C | 70% | MEDIUM |
130| Smoke only | 60% | MEDIUM |
131| Hotspot 60-80°C | 50% | LOW |
132
133### Roof Damage Severity
134| Type | Low | Medium | High | Critical |
135|------|-----|--------|------|----------|
136| Missing shingle | - | - | Always | - |
137| Crack | <1" | 1-3" | >3" | Multiple |
138| Granule loss | <10% | 10-30% | >30% | - |
139| Ponding | - | Small | Large | Active leak |
140
141### Wildfire Risk Factors (Weighted)
142| Factor | Weight | High Risk Indicators |
143|--------|--------|---------------------|
144| Defensible space | 20% | Non-compliant zones |
145| Vegetation density | 20% | NDVI >0.6, high fuel load |
146| Slope | 15% | >30% grade |
147| Roof material | 10% | Wood shake, Class C |
148| Structure spacing | 10% | <30ft between buildings |
149| Access/egress | 10% | Single road, narrow |
150
151### 3DGS Quality Settings
152| Quality Level | Iterations | Time | Use Case |
153|---------------|------------|------|----------|
154| Preview | 7K | 5 min | Quick check |
155| Standard | 30K | 30 min | General use |
156| High | 50K | 60 min | Documentation |
157| Inspection | 100K | 3 hrs | Damage measurement |
158
159## Reference Files
160
161Detailed implementations in `references/`:
162- `fire-detection.md` - Multi-modal fire detection, thermal cameras, progression tracking
163- `roof-inspection.md` - Damage detection, thermal analysis, material classification
164- `insurance-risk-assessment.md` - Hail damage, wildfire risk, catastrophe modeling, reinsurance
165- `gaussian-splatting-3d.md` - COLMAP pipeline, 3DGS training, inspection measurements
166
167## Integration Points
168
169- **drone-cv-expert**: Flight control, navigation, general CV algorithms
170- **metal-shader-expert**: GPU-accelerated 3DGS rendering
171- **collage-layout-expert**: Visual report composition
172- **clip-aware-embeddings**: Material/damage classification assistance
173
174## Insurance Workflow
175
176```
1771. Pre-Event Assessment (Underwriting)
178 ├─ Satellite: Regional risk context
179 ├─ Drone: Property-level risk factors
180 └─ Output: Risk score, premium factors
181
1822. Post-Event Inspection (Claims)
183 ├─ Drone survey: Damage documentation
184 ├─ 3DGS: Measurements, change detection
185 └─ Output: Claim package, cost estimate
186
1873. Portfolio Risk (Reinsurance)
188 ├─ Aggregate: TIV, loss curves
189 ├─ Model: AAL, PML, concentration
190 └─ Output: Treaty pricing, structure
191```
192
193---
194
195**Key Principle**: Inspection accuracy depends on multi-source data fusion. Single-sensor assessments miss critical context. Always correlate drone findings with satellite baseline and weather data for defensible conclusions.
curiositech