1. Analog Control Recognition (Easy)
Analog control recognition involves using AI to automatically read analog gauges, dials, and meters—common in older infrastructure. These readings can be automatically captured and integrated into digital systems for monitoring and analysis.
- Difficulty of Image Acquisition: Low. Analog controls are stationary, easy to access, and generally located in well-lit environments, making it simple to capture high-quality images with standard cameras or existing security footage.
- Modeling Complexity: Low. The task of recognizing and interpreting analog displays can often be achieved using optical character recognition (OCR) or pattern recognition, which are well-established technologies.
- User Acceptance of Automated Solution: High. Many utility operators already use some form of digital monitoring for analog controls, so transitioning to AI-assisted monitoring is generally well-accepted.
Case Study: Utility Alpha automated analog control reading in substations, reducing manual site visits by 70% and improving data accuracy.
2. Image Quality Assessment (Easy)
Image quality assessment ensures that captured visuals are sufficient for their intended purpose by checking images for focus, obstructions, lighting, resolution, and contrast.
- Difficulty of Image Acquisition: Low. Image quality assessment can be applied to images captured by existing cameras or drones, without requiring specialized equipment.
- Modeling Complexity: Low. The models used for image quality assessment analyze straightforward visual attributes like sharpness, brightness, and noise levels, which do not require advanced AI techniques.
- User Acceptance of Automated Solution: High. Automating image quality checks ensures that all images meet necessary standards while preventing return visits to capture better images.
Case Study: Utility Beta automated detection of blurry drone inspection images of wind turbines, reducing the need for return visits by 80%.
3. Object Detection (Moderate)
Object detection identifies objects within captured images, such as equipment, vehicles, or personnel in utility facilities, and can monitor critical infrastructure or ensure safety protocols are followed.
- Difficulty of Image Acquisition: Moderate. Fixed cameras face few challenges, but ground inspections may have obstructions or difficult angles. Aerial image capture requires drones or helicopters with trained operators.
- Modeling Complexity: Moderate. AI models must be trained to recognize specific objects in diverse environments, accounting for differences in size, shape, orientation and background.
- User Acceptance of Automated Solution: Moderate to High. Users are already familiar with object detection through everyday applications.
Case Study: Utility Delta automated identification of bird nests on transmission towers using drone imagery, saving hundreds of labor hours.
4. Defect Detection (Moderate/Difficult)
Defect detection aims to identify issues such as cracks, corrosion, or other signs of wear on equipment, or find if parts that should be present are missing.
- Difficulty of Image Acquisition: Moderate/High. Acquiring images for hard-to-reach equipment requires drones or specialized cameras and may be impacted by environmental factors.
- Modeling Complexity: Moderate. Detecting defects requires AI models to accurately identify subtle signs of wear or damage while differentiating between normal variations and critical defects.
- User Acceptance of Automated Solution: Moderate. Professional inspectors are cautious about fully relying on AI. A human-in-the-loop approach allows them to build confidence.
Case Study: Utility Gamma automated detection of missing insulator discs on transmission lines using drones and AI, improving maintenance response times and grid reliability.
5. Distance Assessment (Moderate/Difficult)
Distance assessment evaluates whether trees, vegetation, or other objects are too close to critical electrical infrastructure for safety or regulatory purposes.
- Difficulty of Image Acquisition: Difficult. Acquiring accurate images in large, remote, or densely forested areas is challenging. Drones, helicopters, or satellite imagery are often used.
- Modeling Complexity: High. AI models need to accurately assess distances from 2D images, involving depth estimation and understanding of scale in various environments.
- User Acceptance of Automated Solution: Moderate. Utilities increasingly accept automation for straightforward cases, especially when AI can output "cannot determine" for unclear images.
Case Study: Utility Epsilon automated vegetation encroachment assessment near transmission lines, predicting when trimming would be needed based on growth rate analysis.
6. Orientation Assessment (Difficult)
Orientation assessment determines the exact position and alignment of objects, critical for ensuring proper installation or detecting misalignments in infrastructure.
- Difficulty of Image Acquisition: Difficult. Capturing necessary images requires precision, often from multiple angles, and may require specialized equipment or controlled conditions.
- Modeling Complexity: Very High. Orientation models must interpret 3D space and rotational positioning of objects.
- User Acceptance of Automated Solution: Low to Moderate. Due to complexity, users may hesitate to trust fully-automated solutions when misalignment could have costly or dangerous consequences.
Case Study: Utility Zeta partnered with street-level vehicles to automate detection of leaning utility poles, reducing dedicated image capture needs.
Where to Begin
When determining where to begin with Visual AI, it's essential to consider your organization's current capabilities, resources, and goals. For utilities starting with AI, beginning with easier use cases like analog recognition or image quality assessment delivers quick wins and builds confidence. As teams become more familiar, they can progress to moderate or difficult use cases where operational impact is higher but more challenging to achieve.