Mobility Sensing and Analytics for Smart Cities
REU Mentor: Jiannan Zhai, Ph.D. & Jason Hallstrom, Ph.D.
REU Scholar: Kade Townsend
REU Scholar Home Institution: Southwestern University
Mobility Sensing and Data Analytics for Smart Cities
Understanding mobility patterns of people is an essential part of the economic development and optimization of needed services within a city. This can be done with MobIntel, a privacy-centric alternative to surveillance using signals from WiFi-enabled devices. In this project, we use the ideas of descriptive analytics and trendline forecasting to detail and verify the data readings of our sensors. This is a relatively new product, so our results have to be fact-checked. The Seaborn and Matplotlib libraries of the Python programming language are used to create many different visual representations of how the signal counts and different variables relate to each other. We then analyze these illustrations to make sense of the mobility patterns within the area and validate our findings with other ground-true data. Furthermore, the sensors are subject to loss of power for various reasons. Therefore, a Python-based program utilizes past data and trendlines to find the power status of our sensors in the field. In continuing with this undertaking, we hope to use more sophisticated machine learning in order to create a model that can predict the number of device counts for any given timeframe.