This is a past event.
Thursday, October 1, 2020 4:00pm
This is a past event.
Focusing on the Problem of Occupant Localization and Path Tracking using Hidden Floor Accelerometers
ME-EM Virtual Graduate Seminar Speaker Series
Sa’ed Alajlouni, PhD
Abstract: A living environment can be made smarter by measuring floor vibration, and extracting human-related activity information from the vibration measurements. Sensors that measure floor vibration include strain gauges (measuring floor displacement at the point of installation), geophones (measuring velocity), and accelerometers (measuring acceleration).
An underfloor network of hidden accelerometers provides a passive and tamper-proof system for monitoring human activity. The system is non-intrusive compared to a camera-based monitoring system, and does not require occupant compliance by wearing an electronic device (i.e., device-free).
Human activity monitoring includes: 1) event detection & classification (e.g., does the detected vibration event corresponds to a footfall? a person falling and hitting the floor? a gunshot? a door opening/closing? a chair moving? etc.); 2) event localization (e.g., estimating the location of an identified footfall (footstep)); and 3) inference (e.g., estimating the number of people in a room, and analyzing gait style or speed).
In this talk, the speaker will focus on the problem of occupant footstep impact localization and occupant path tracking, using floor accelerometers. The talk will include: explaining different approaches of solving the general source localization problem; showing the challenges in localizing floor impacts including “wave dispersion”; and motivating the development of energy-based (received signal strength (RSS) based) footstep localization methods.
Bio: Sa’ed Alajlouni is an assistant professor in the department of Mechatronics Engineering at the Hashemite University (Jordan; middle east). His research interests include: impact localization, input-force estimation (inverse problems), signal processing, optimization, algorithm development, and vibration data analytics in general. He currently teaches topics including “Signals and Systems”, “Building Automation Systems”, “Instrumentation”, and “Controls”. Dr. Alajlouni has a PhD in Electrical Engineering from Virginia Tech (focus: signals, systems, and controls); a MSc in Electrical Engineering from Texas Tech University; and a BSc in Mechatronics Engineering from the Hashemite University. In addition to his education, Dr. Alajlouni has industry experience, where at some point during his career, he worked as a freelancer in the field of industrial process control design and maintenance, with an intensive experience in the process of “plastic injection molding”. Also, at a different point in his career, Dr. Alajlouni worked in radioactive waste management.
Dr. Alajlouni’s early publications were in the fields of “laser beam path tracking, using Risley prisms” and “anomaly detection in cyber-physical systems”. He has a number of publications extending from his PhD dissertation, in which he focused on analyzing big vibration data from human activity inside the world’s most instrumented smart building for vibrations measurements; Goodwin Hall (Virginia Tech). He mainly used vibration measurements, from underfloor accelerometer sensor networks, to locate footsteps and track occupants. His research was interdisciplinary; consisting of theoretical, experimental, and simulation parts, including physics-based modeling, wave propagation in solids, instrumentation and data acquisition, signal processing, stochastic processes, and feature extraction.
Invited by: Sriram Malladi