AI Use Cases

The platform addresses all 17 mandatory AI use cases from the RFP.

Use Case Summary

#	Use Case	Primary Sensor	AI Approach
1	Fluid detection on ground	Thermal + Visual	Pattern recognition
2	Flow detection in lines/wells	Thermal	Thermal signature analysis
3	Liquid/gas leak detection	Thermal + Methane	Multi-sensor correlation
4	Methane (CH4) detection	Methane (TDLAS)	Sensor + wind compensation
5	Medium voltage line inspection	Visual + Thermal	Anomaly detection
6	Construction progress	LiDAR + Visual	Change detection
7	Unauthorized objects	Visual	Object detection
8	Liquid levels in chemical skids	Visual	Level recognition
9	Person/vehicle intrusion	Visual	Object detection + tracking
10	Volumetric calculation	LiDAR	Point cloud processing
11	Hot spots in equipment	Thermal	Temperature anomaly
12	Road/access anomalies	Visual + LiDAR	Surface analysis
13	Electromechanical protection	Visual	State recognition
14	Perimeter fence integrity	Visual	Damage detection
15	Mechanical integrity	Visual + Thermal	Condition assessment
16	Valve/tank operational status	Visual	State classification
17	Underground interference (GPR)	GPR	Assisted interpretation

Detection Categories

Leak Detection (Use Cases 1-4)

Multi-sensor approach for comprehensive leak detection:

┌─────────────┐    ┌─────────────┐    ┌─────────────┐
│   Visual    │    │   Thermal   │    │   Methane   │
│   Pattern   │    │  Signature  │    │   Reading   │
└──────┬──────┘    └──────┬──────┘    └──────┬──────┘
       │                  │                  │
       └──────────────────┼──────────────────┘
                          │
                ┌─────────▼─────────┐
                │    Correlated     │
                │   Leak Detection  │
                └───────────────────┘

Security (Use Cases 7, 9, 14)

Object detection and tracking for security monitoring:

• Person detection with tracking
• Vehicle classification
• Unauthorized object identification
• Perimeter breach detection

Asset Integrity (Use Cases 5, 11, 13, 15, 16)

Visual and thermal analysis for equipment status:

• Thermal anomaly patterns
• Visual condition assessment
• State recognition (open/closed)
• Degradation indicators

Measurement (Use Cases 6, 8, 10, 12)

Quantitative analysis capabilities:

• Volumetric calculation from LiDAR
• Liquid level estimation
• Construction progress tracking
• Surface condition analysis

Subsurface (Use Case 17)

GPR-based underground detection:

• Assisted interpretation
• Anomaly visualization
• Expert review workflow

Accuracy Targets

Category	Target	Measurement
Detection rate	≥95%	True positive rate
False positive	<10%	False alarm rate
Localization	±10m	GPS accuracy
Classification	≥90%	Correct category

Phased Implementation

Phase	Use Cases	Focus
Phase 1	Data collection	Baseline for all 17
Phase 2	1-4, 7, 9, 11	Core detection
Phase 3	5, 13-16	Asset integrity
Phase 4	6, 8, 10, 12, 17	Measurement + GPR