Projects Funded for Mehdi Nemati


The Impacts of Wildfires on Water Utilities and Communities in California

  • Samane Zare
  • Mehdi Nemati


Unpacking Residential Water Consumption and the Impacts of Nudges: A Machine Learning Application

  • Mehdi Nemati


Specific Objectives of the Project:

  • Disaggregate residential water consumption to indoor and outdoor usage using machine learning methods.
  • Estimate effect of HWURs on indoor and outdoor water consumption (obtained in the first objective).
  • Estimate the impact of HWURs on peak hour and day water consumption.

Project Progress Report:
Using around 600 million hourly water use data and machine learning methods, including Random Forests (RFs), Artificial Neural Networks (ANNs), and Support Vector Regression (SVR) we were able to disaggregate total water use to indoor and outdoor usage. The preliminary results indicate that 60% of total water use is related to outdoor activities, mainly lawn irrigation, and happens between 10 pm and 4 am. The pick hour consumption is between 12 am and 3 am.
For the two other objectives, we merged the disaggregated data with census demographics and weather information such as temperature and precipitation. In addition, we observe Home Water Use Repots message content and timing for the enrolled households. Currently, we are in process of generating and processing the models for the second and third objective.