Predicting failure & automating defect detection on utility assets
summary
PIPE AI deploys a cloud pipeline on AWS (ECS, EFS, Lambda, Direct Connect/PrivateLink) that automates CCTV review with computer vision and a Review and Predict split: Review flags defects, Predict scores assets and forecasts maintenance. A recent evaluation finds 313 vs 103 valid detections (human v. model) across sample videos, with nine videos yielding 12 false positives and 3 false negatives; system reports up to 15% more defects and offers one‑click inspections, retraining via annotated reviews and a communal model.
AI-written summary · about
chapters (5)
- 0:00Start
- 7:32Benefits
- 18:20Model analysis (recent example)
- 21:59Challenges and lessons learned
- 23:34Current plans and future visions
related talks
36mAI-Powered Real-Time Quality Control
29mAI-Powered Anomaly Detection in Cloud Data Pipelines
22mEngineering Predictive Platforms
29mAmazon Lookout for Vision
21mAI-driven optimization of CI/CD pipelines
26mServerless Architecture for Product Defect Detection
30mIntelligent Pipelines: Bringing AI/ML into DevSecOps Automation
05mAutomating Cloud-Native Infra with Intelligent Tools