Use cases
Transportation infrastructure — roads, bridges, rail, overhead power lines, port cranes — degrades invisibly until it fails. Computer vision applied to drone and fixed-camera imagery catches defects before they become disasters.
Road & Rail Infrastructure Inspection
Drone imagery run through object detection models identifies surface cracks, track deformations, overhead line anomalies, and signage damage — without requiring track or road closures.
Cargo & Load Image Quality Validation
Before shipping documentation is generated, an AI quality gate validates that load photos are clear, correctly angled, and unoccluded — the same pipeline ThirdEye Data built for Southern California Edison's electric pole inspection.
Anomaly Detection on Electric / Rail Infrastructure
Detect defective components, predict deterioration probability, and route anomalous assets to maintenance queues — directly adapted from ThirdEye Data's Microsoft anomaly detection work on electric poles.
Automobile & Freight Quality Check (Sound + Vision)
ThirdEye Data's CenturyPly project detects internal defects by analysing audio signatures — the same approach applies to rail track tap-testing and structural integrity checks at freight depots.
ThirdEye Data capabilities mapped
Computer Vision (TensorFlow)
Object Detection (YOLO)
Structure Detection
Occlusion / Blur / Angle AI
Drone Image Pipelines
Anomaly Detection
Audio / Spectrogram ML
Master Image Pipeline
Two anchoring reference projects
SCE Image Quality Detection (Southern California Edison): End-to-end platform ingesting third-party images of electric poles, detecting structure, tags, occlusion, blur, and angle — then routing quality-cleared images to downstream anomaly detection. Stack: TensorFlow, Google Vertex AI, Azure Data Factory, CosmosDB.
Microsoft Anomaly Detection in Electric Poles: Drone-captured images ingested to Azure Blob; AI identifies poles, detects anomalies in sub-parts, predicts deterioration probability. Used by Microsoft's customers and field sales worldwide.