Overview:
Exelon partnered with Deloitte and NVIDIA to develop OptoAI, an autonomous drone inspection solution that uses AI and edge computing to enhance grid infrastructure inspection. The system aims to replace manual and pilot-guided drone inspections by enabling fully autonomous flight and real-time AI-driven analysis of power lines, poles, and other grid assets. The solution operates within utility field operations to increase efficiency, reduce risk, and accelerate maintenance decisions for infrastructure teams responsible for grid reliability and safety.
Key Features:
- Autonomously controls drone missions with built-in AI models to locate and inspect grid assets without constant human piloting.
- Runs computer vision models on the edge using NVIDIA Jetson modules to detect defects like broken crossarms and leaning poles in real time.
- Generates synthetic training data via NVIDIA Omniverse to expand model robustness for various inspection scenarios.
- Integrates with geographic information systems (GIS) to auto-locate assets and correct location data during inspection flights.
- Delivers prioritized inspection results instantly to field crews via web, tablet, or extended reality interfaces for actionable insights.
Results & Impact:
- Achieved over a 100× reduction in inspection planning and operation time, compressing some missions from ~60 minutes to ~30 seconds.
- Reduced human risk and error by minimizing direct exposure of field workers to hazardous inspection environments.
- Turned weeks of manual image sorting and analysis into seconds of instantly prioritized insight for maintenance crews.
- Enhanced overall grid reliability by accelerating defect identification and enabling proactive, real-time response.
- Received industry recognition, including a Public Utilities Fortnightly Edison Pioneers Innovator Award for collaboration in innovation.
AI Technology:
AI Model Types: Computer vision, synthetic data augmentation
AI Purpose: Automate inspection and defect detection, enable edge vision analytics
Application Type: Energy operations, field asset management
Target Users:
- Grid repair and maintenance crews
- Utility field operations teams
- Asset reliability and inspection specialists
- Innovation and technology leaders within energy utilities
Sources:
- nvidia.com/en-us/customer-stories/exelon