Master's Project

Ground Work for the Master's project

Applying Machine Learning and Crowdsourcing in Utility Management Services

Project Contributors

Name	Email
Meera Tresa Sebastian	meeratresa.sebastian@sjsu.edu
Adesh Landge	adesh.landge@sjsu.edu
Vineeth Hamilpur	vineeth.hamilpur@sjsu.edu
Chiruhas Bobbadi	chiruhas.bobbadi@sjsu.edu

Automated Meter Reading, Software Engineering, Image Processing, Deep Learning, Human-in-the-Loop

Human-in-the-loop image-based Automated Meter Reading System

We are developing a comprehensive, cloud-based automation system designed for non-smart meters, utilizing deep learning-based image recognition techniques and combining it with a human-in-the-loop verification framework based on crowd-sourcing and continuous retraining. We hope to create a robust and accurate solution for automated meter reading from customer-uploaded images of utility meters. Our vision is to supplant the currently expensive, labor-intensive, and error-prone process of manual operator-based reporting with self-reporting of meter readings for utility management companies. By doing this, we can enhance the efficiency of utility management. This also offers a viable solution for developing nations to maximize the use of their existing non-smart utility meters infrastructure.

Shared Drive