Dissertation Proposal - University of Houston
Skip to main content

Dissertation Proposal

In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

Nacer Khalil

will defend his dissertation proposal

Occupant Identification and Navigation Using Ultrasonic Sensing


The ability to non-intrusively identify occupants enables solving many problems in smart buildings such as the need to personalize the climate which would save energy and enhance occupants’ comfort. To achieve that, we propose a method to identify occupants by sensing their body shape and movement as they walk through the door. We mount three ultrasonic ping sensors, one on top to sense height and two on the sides of the door to sense width. We extract a feature set from the occupant’s walk to non-intrusively identify him. We cluster the occupants using their waist girth and the time spent under the door. We use DBSCAN as for clustering because it discovers the number of clusters and also takes into consideration the precision of the sensors. We deployed our system in a classroom for a month, and 20 people participated. Our model is able to identify 20 occupants with an accuracy of 95%. This technology could be used in various applications but larger identification population would be required. To reach a population size of over 100 people, we need to improve on different components of our system. First, we increase the sampling rate from 30Hz to over 145Hz using sensor sampling optimization and height width sampling parallelization. We also increase the precision of the sensors by reducing the error from 1cm to 0.3cm by optimizing the intensity and length of the sound beams used to measure the distance. We deploy five door frames over an extended period of time to collect a larger dataset with higher data precision. A larger and more fine grain dataset enables more complex feature explorations which would achieve larger identification populations at about 95% accuracy. In addition, a larger dataset makes it possible to explore various deep learning models. Increasing the population could be leveraged by different applications targeting occupants indoor. In fact, device-free indoor navigation would leverage such a technology. It consists of guiding users to reach their destinations in large commercial buildings. In many cases, some areas in buildings could be more congested than others. Our technology by being non-intrusive and device-free can help guide the users by identifying and re-identifying them.

Date: Wednesday, February 22, 2017
Time: 10:00 AM
Place: PGH 550
Advisor: Prof. Omprakash Gnawali

Faculty, students, and the general public are invited.