Projects and Research

>Faculty Involved: Dr. Yong-Kyu Jung

This research explores information security & integrity in cyberspace, cyber-physical systems, rapid prototyping with FPGA, and real-time simulator with various approaches to heterogeneous modeling and prototyping.

 

Faculty Involved: Dr. Wookwon Lee

This research project is to develop a complete near-space ballooning system in preparation for the spectacular natural event of total solar eclipse on April 8, 2024 (more specifically, full beginning at 3:16:22 pm EDT, maximum occurring at 3:18:13 pm, and full ending 3:20:05 pm). The passage of the 2024 total solar eclipse includes the skies right above the downtown of Erie, PA and the shoreline of Lake Erie (right at the home ground of Gannon!).

Faculty Involved: Dr. Wookwon Lee

This research is to develop a science payload to detect cosmic rays, high-energy particles of astrophysical origin, in the energy range ~1–100 GeV, and also create an analytical model that relates the energy of cosmic rays to the output of an electronic integrator. The payload will be carried to an altitude of 120,000 ft for a flight duration of ~6 hours on a Small Balloon System operating in close collaboration with NASA’s Balloon Program Office.

Faculty Involved: Prof. Don MacKellar

This project is for blended engineering product development and also for participation in professional competition of autonomous systems. IGVs are aimed to be able to recognize, avoid, and navigate through dynamic obstacles. Artificial intelligence, machine learning, and control techniques are explored.

Faculty Involved: Dr. Fong Mak

This project is to implement a Person Identification and Health (Temperature) Scanned System that replaces the existing manual attendance and daily thermal recording requirement for K-12 schools or institutions to improve the current manual or COVID-monitoring processes.

K12 schools have been manually tracked their students' daily attendance records by taking counts of the students when they are either in a classroom or entrance to a school.  In addition, some schools' policy requires schools to take students' daily temperature for their health-tracking requirements.  The existing systems either have these two attendance and temperature scans, processes conducted separately, or lack a coordinated database system for these two processes.  Coupled with the COVID-19 situation, the need to have a better system becomes urgent. This project is supported by MAKTEAM Software.

The significant challenges for the project are (1) integrating embedded hardware, web technology, and cybersecurity knowledge into this product development, (2) security authentication of each endpoint, (3) constant health check of each endpoint, (4) secure communication between endpoint and command center, (5) integration of collected data with the existing SIS, (6) capable of performing Q.R. code or facial identification, (7) temperature scanned is accurate in all in-door environmental conditions, and (8) dashboard design that appeals to customers.

Faculty Involved: Dr. Ramakrishnan Sundaram

Radio frequency signals can be used to perform non-invasive and device-free target localization of objects or entities in space. Radio tomographic imaging uses wireless sensor networks to form images from the attenuation of the radio frequency signals. Radio tomographic imaging is useful to locate security breaches, to perform rescue operations, and to design “smart” buildings. The integrated radio tomographic imaging system comprises subsystems identified as the wireless sensor network, the command and data collection platform, and the user interface. The project will set up the space hardware laboratory to assemble and test the subsystems, to design and document the project activities on the integrated radio tomographic imaging system, and to deliver STEM outreach with pK-12 schools using the integrated system.

Faculty Involved: Dr. Ramakrishnan Sundaram

This research is to enable the general public to have better communication with hearing-impaired people through the American Sign Language by using an automated translation system. The design of the translation system is based on the current state-of-the-art machine-learning technology, specifically the Deep Convolutional Neural Networks, which is one of the sub-fields of Artificial Intelligence. The models designed and tested as part of this research, performed significantly better and learned faster when mask videos were used instead of raw videos, and fusion nets performed the best, with the best per-class top-1 accuracy of 96.6% over the 500-class dataset. More importantly, it confirms that keeping a constant temporal dimension in a deep network is a viable approach to sign language recognition, especially when the temporal information from all the available snapshots can be fully exploited by a time-sensitive construct such as ConvLSTM.