Files and links (1)
Ghadimzadeh-Alamdari_Ali_2025152.65 MB
PDF Embargoed Access, Embargo ends: 31 Dec 2026
Dissertation
A multi-scale and multi-sensor data fusion and digital twin framework for artificial intelligence- and robotic-driven inspection of civil infrastructure
Doctor of Philosophy (Ph.D.), Drexel University
Dec 2025
DOI:
https://doi.org/10.17918/00011234
Abstract
The assessment and maintenance of civil infrastructure have traditionally relied on manual, experience-driven inspection procedures that are subjective and present significant safety risks and operational hazards. This dissertation introduces a robotics and artificial intelligence (AI)-based framework that aims to transition conventional inspection practices toward a scalable, automated, and repeatable procedure. Followed by a review and benchmarking for simultaneous localization and mapping (SLAM) techniques for environmental perception, this dissertation implements a decision-making process inspired by human inspectors. The proposed approach begins with a broad contextual understanding of the environment and incrementally narrows focus to regions requiring detailed evaluation. At each level of inspection a defect type was inspected and quantified during the robotic inspection. At the large scale, structural warping is detected and quantified using light detection and ranging (LiDAR), with validation conducted both in simulation and on a salvaged girder from a real-world interstate highway bridge. At the surface scale, camera vision combined with AI identifies regions of interest associated with cracks, which are then examined using a robotic arm equipped with a high-resolution laser surface profiler for detailed geometric characterization. At the subsurface scale, normal estimation from segmented point clouds guides robotic placement of sensors on inclined and curved surfaces, that enables targeted ultrasonic pulse velocity (UPV) measurements to detect delaminations near the surface and assess through-thickness integrity of non-flat concrete structures, including those arising in additive manufacturing. Finally, a remote inspection workflow through a digital-twin and virtual-reality interface allows off-site experts to interact with the inspection scene in real time. The proposed framework demonstrates a shift in paradigm by integrating localization, multi-scale sensing, contact-based nondestructive evaluation (NDE), and remote accessibility, thereby enhancing the safety, consistency, and scalability of inspection operations.
Metrics
28 Record Views
Details
- Title
- A multi-scale and multi-sensor data fusion and digital twin framework for artificial intelligence- and robotic-driven inspection of civil infrastructure
- Creators
- Ali Ghadimzadeh Alamdari
- Contributors
- Arvin Ebrahimkhanlou (Advisor)Ivan Bartoli (Advisor)
- Awarding Institution
- Drexel University
- Degree Awarded
- Doctor of Philosophy (Ph.D.)
- Publisher
- Drexel University
- Number of pages
- xvi, 232 pages
- Resource Type
- Dissertation
- Language
- English
- Academic Unit
- College of Engineering (1970-2026); Mechanical Engineering (and Mechanics) [Historical]; Drexel University
- Other Identifier
- 991022147107004721