Using Wifi-Based Occupancy Data to Increase Energy Efficiency for Smart Buildings


  • Rithvik Gujjula Aspiring Scientists' Summer Internship Program Intern
  • Dominic King Aspiring Scientists' Summer Internship Program Intern
  • Dr. Duminda Wijesekera Aspiring Scientists' Summer Internship Program Mentor
  • Dr. Kamaljeet Sanghera Aspiring Scientists' Summer Internship Program Mentor



Although the efficiency of HVAC systems has improved over the years especially for “smart buildings”, there hasn’t been significant improvement in terms of maximizing the efficiency of energy output throughout the day in a building. Often, HVAC systems have pre-programmed cycles which are excellent for residential settings - in which people will spend more time, especially due to the pandemic - but not necessarily for commercial buildings. With climate change already showing its effects on the environment, it’s clear that systems that are able to reduce energy wastage will be critical.

The primary purpose of the research is to identify how to use real-time wifi-based occupancy data in order to increase energy efficiency. Over the past few weeks, our team has designed a script to help identify the number of devices and consequently the number of people in a certain room of a building using a machine learning model. The script retrieves input from wifi logs such as the signal strength and the number of connected devices. In addition, the script uses the timestamps to determine at which time of the day most devices are connected. The number of devices along with the timestamp is fed into a regression model to predict the occupancy.

The ramifications of technology that can maximize or reduce energy output based on real-time data on the number of people in a building are vast. Eventually, the goal is to build an HVAC system that can rapidly adjust its output based on real-time occupancy data and reduce energy wastage.





Institute for Digital Innovation