Evolution of Compliance Risk Assessment

February 05, 2026
10
Storms and Wildfires , Asset Management , Policy and Regulation

Risk assessment is a top priority for SDG&E as wildfire threats to California communities continue to intensify. Events like the Palisades and Eaton fires underscore the scale of the challenge. One of our most impactful initiatives has reimagined compliance by integrating computer vision and advanced risk modeling to enhance situational awareness and enable proactive management, ultimately delivering safer, more reliable power to the communities we serve. Through modernization of legacy inspection processes, the Digital Innovation program is enhancing safety, operational efficiency, and regulatory adherence across its service territory.  

To maximize the value of visual data, the Digital Innovation team developed the Intelligent Image Processing (IIP) platform. IIP uses artificial intelligence, computer vision, and machine learning to analyze imagery from drones, field crews, and other sources. It automates the detection of infrastructure issues, vegetation encroachment, and asset degradation, significantly reducing manual review time. By transforming images into actionable insights, IIP accelerates decision-making, supports predictive maintenance, and enhances situational awareness. For key use cases like field technicians uploading site photos, IIP ensures that critical visual information is no longer siloed or underutilized. Instead, it becomes a high-value data stream that feeds predictive maintenance models and supports compliance and risk mitigation efforts. 

In 2021, SDG&E’s Digital Innovation team launched a proof of concept to improve real-time risk visibility during Public Safety Power Shutoff (PSPS) and wildfire events. This effort led to a major enhancement of the Wildfire Next Generation System (WiNGS) with the Temporary Construction and Compliance (TCC) feature. TCC provides a geospatial view of over 7,000 temporary construction poles, integrated with real-time data from circuits, upstream/downstream devices, and weather stations. During a recent Emergency Operations Center (EOC) activation, this capability enabled proactive field mitigation, contributing to a 50% reduction in PSPS events, minimizing outages and improving community safety.  

SDG&E began the Risk-Informed Drone Inspection (RIDI) initiative in 2023 to leverage computer vision and risk models with new drone capabilities. RIDI conducts approximately 13,000 pole inspections annually using drones equipped with high-resolution imaging. These drones capture critical asset details (like pole tops and crossarms) and access remote, rugged terrain without exposing crews to hazardous conditions. The data can then be ​uploaded and analyzed within the IIP UI for analysis. The future of damage detection will also integrate synthetic data, using SD to replicate highly unlikely but very bad damage that the models are not trained for. 

Together, these programs exemplify SDG&E’s commitment to innovation, safety, and operational excellence. By integrating drones, AI, geospatial analytics, and machine learning into core utility functions, the business is setting a new standard for how utilities can modernize infrastructure management, enhance compliance, and protect the communities they serve.  

Speakers
Cameron Carroll
Cameron Carroll, Group Product Manager Digital & Portfolio Management - San Diego Gas & Electric
Chairperson
Abder Elandaloussi
Abder Elandaloussi - Southern California Edison