SoPhySens: A Novel Sensing Paradigm That Explores the Collective Intelligence of Humans and Machines
Propelled by the coexistence of diverse data capture, communication, and computing technologies, physical sensing has revolutionized the avenue for accurately capturing real-world phenomena. Despite their virtues, various limitations hinder their efficacy in critical scenarios such as disaster response. On the other hand, social sensing is contriving as a pervasive sensing paradigm that leverages observations from human “sensors” to perceive the environment (i.e., through social media or crowdsensing). While social sensing introduces many benefits that address some caveats of physical sensing, social sensing also inherently suffers from a few drawbacks: inconsistent reliability, uncertain data provenance, and limited sensing availability. Motivated by the reciprocal virtues of social and physical sensing, we focus on the concept of social-physical sensing (SoPhySens) in this dissertation. This novel integrated sensing paradigm unifies human wisdom from social sensing with the empirical sensing prowess of physical sensors to reconstruct the state of the world, essentially drawing a complete picture of real-world occurrences that might not be possible with standalone sensors.
For the scope of this dissertation, we intend to address three fundamental challenges in SoPhySens, specifically: i) how to reliably acquire relevant raw signals from multitudes of social and physical sensors and relate the collected data to each other Md Tahmid Rashid given their diverse characteristics? ii) How to efficiently handle the complex interactions between the human, cyber, and physical components in SoPhySens when melding social and physical sensing? iii) How to adapt to the intricate dynamics that arise when jointly exploring the physical world and the social domain? To address these challenges, we developed multi-dimensional analytic frameworks and prototypes. As our foundational work, we present a novel closed-loop social-physical active sensing infrastructure called social airborne sensing (SAS) that leverages social media signals to locate events of interest during disaster scenarios and dispatches unmanned aerial vehicles (UAVs) to investigate the reported locations. Afterward, we expand our SAS infrastructure to existing vehicular infrastructure by introducing a social media-driven car sensing system, namely social vehicular sensor networks (S- VSN), for scalable post-disaster recovery. Lastly, we present Chirper, a collaborative platform that enables the interplay of UAVs and humans to coordinate situational awareness tasks in crisis response scenarios, specifically search and rescue (SAR). The developed solutions promise the groundwork for holistic situation awareness that can potentially unravel episodes from the real world and significantly extend the current landscape of SoPhySens from both analytic and system perspectives.
History
Defense Date
2023-07-28CIP Code
- 40.0501
Research Director(s)
Jane Cleland-HuangDegree
- Doctor of Philosophy
Degree Level
- Doctoral Dissertation
OCLC Number
1413472423Additional Groups
- Computer Science and Engineering
Program Name
- Computer Science and Engineering