Learn more about innovative approaches and use AI to optimize operations and reduce costs in the utility sector.
In the utility sector, optimizing operations and reducing costs are critical goals. One pressing challenge is determining whether gas meters that register zero consumption need to be investigated, which traditionally requires costly and labor-intensive manual reviews. For a mid-size utility, this process not only involved sending out “truck rolls” but also maintaining a team of eight full-time employees (approved to scale department to 16 FTEs) to review zero consumption reads.
To address this challenge, our team embarked on an innovative project to streamline the process using Narrow AI. We began by studying the monthly meter read process to understand its intricacies and pain points. Our next step was to integrate the monthly meter read system with customer billing history and customer service systems, creating a comprehensive data set that includes customer usage patterns, weather conditions, and historical data.
With this robust data foundation, we developed a predictive model designed to analyze and identify patterns indicating whether a gas meter is likely failing. This model uses advanced machine learning algorithms to make predictions based on the aggregated data, providing actionable insights with remarkable accuracy.
The results were outstanding. Our predictive model achieved a 98% accuracy rate in determining failing meters and the need for investigation. This high level of precision enabled the utility to significantly reduce the labor required for manual meter read reviews, effectively allowing the reallocation of human resources to more critical tasks. Additionally, the reduction in unnecessary truck rolls led to substantial cost savings, cutting down on fuel expenses and vehicle wear and tear.
The implementation of this Narrow AI solution has transformed the utility’s approach to meter maintenance. By leveraging AI to predict failing meters, the utility can proactively address issues, ensuring continuous service and enhancing customer satisfaction. This project not only demonstrates the practical benefits of AI in the utility sector but also sets a new standard for operational efficiency and cost-effectiveness.
Written By:
Cody Krietemeyer, Practice Leader, Data Engineering & Data Science
Data Strategy Group
