Evidence Package

RFP Compliance: Section 6.2 Proof of AI Use Cases

This section demonstrates Trifork's capability to deliver the 17 mandatory AI use cases through production deployments and technical proof points.

Section	Description
Rail Yard Case Study	Primary evidence - identical hardware stack
Semco Maritime	Corrosion detection in industrial environments
VELUX	Manufacturing defect detection

Evidence Summary

Trifork's autonomous rail yard monitoring project for a major US maintenance operator provides the strongest evidence of capability:

The following components transfer directly to YPF:

Component	Rail Yard Implementation	YPF Application
Triton Ensemble	GroundingDINO + SAM2 + OCR	Multi-sensor detection
Digital Twin	PostGIS geospatial	Asset correlation
Processing	Micro-batch pipeline	20-min SLA compliance
Clustering	Hybrid algorithm	Multi-signal fusion
OCR	Probabilistic reconciliation	Equipment identification

#	Use Case	Evidence Source
1	Fluid detection	Rail yard visual detection
2	Flow detection	Thermal + visual correlation
3	Liquid/gas leak	Multi-sensor fusion
4	Methane detection	Sensor integration architecture
5	Power line inspection	Visual anomaly detection
6	Construction progress	Volumetric measurement
7	Unauthorized objects	Object detection AI
8	Liquid levels	Visual/thermal analysis
9	Intrusion detection	Person/vehicle detection
10	Volumetric calculation	LiDAR processing pipeline
11	Hot spots	Thermal pattern detection
12	Road anomalies	Visual inspection AI
13	Protection verification	Equipment status detection
14	Fence integrity	Perimeter analysis
15	Mechanical integrity	Visual defect detection
16	Valve/tank status	Component recognition
17	Underground (GPR)	Subsurface data visualization

Note: Visual evidence (screenshots, architecture diagrams, demo videos) to be compiled separately. See PRD visual evidence checklist.

Planned visual assets: