Project Proposals


Smart Campus

Problem Statement: We have the IT infrastructure, yet most of services or functions provided/performed on campus are manual, e.g. taking attendance, paying for food and drinks, … etc. We need a smart campus system where all services and operations are automated. Example services include automatic presentation loading for faculty, automatic note-taking, auto-login to campus computer facilities, auto-charging for services such as cafeteria and photocopying.

Prerequisite Courses: ICS 202, COE 306, COE 344, and COE 444 or COE 405 or COE 424

Required Skills: Unlimited imagniation, HW/SW/Networking proficiency

Outcome: A complete system with at least one provided service.

Project Requirements:

  • Automatic recognition of faculty, staff and students.
  • Customizable service levels.
  • Protected users’ privacy and data security.
  • Non-invasive solution, i.e. does not require users to wear anything!
Advisor: Dr. Muhammad Elrabaa No. of students: 3

Mawjood: An OpenCV-based Embedded System for Automatic Attendance-Taking

Problem Statement: A hall is filled with 1000 students. We would like to know who is present. Taking attendance manually is a very time-consuming process. Each proctor is given a device that he can use to scan the faces of the students in the group for which he is responsible. A central server will then report to the course coordinator the names and IDs of the students who are missing.

Prerequisite Courses: COE 487, COE 306

Required Skills: C/C++

Outcome: Prototype.

Project Requirements:

  • Device with a camera module that detects faces. The student has to decide if face recognition would be possible on the device.
  • Centralized node (can be a cloud-based app, server, laptop, …) to perform recognition (if not done on device) and generate reports for the course coordinator.
Advisor: Dr. Yahya Osais No. of students: 3

Enhancing the Parallel N-Body Simulation Time on Intel MIC

Problem Statement: This project is to enhance the performance of a parallel N-Body Simulation Program written in C with OpenMP directives for Intel MIC, a 60-core Xeon Phi available in the Robotics Lab. The N-Body simulation is an iterative procedure that repeatedly computes the gravitational forces among N bodies which cause the body motion. The brute force approach is (O(N2)). An approximation is applied to achieve O(N log(N)). The parallel program has three steps: (1) building the needed data structure based on knowledge of body positions, (2) parallel computation of the forces, (3) updating the position. About 90% of the parallel time is spend in step 2. The project is to (1) reduce the parallel time by tiling the massive data to enhance data locality (reuse), and (2) optimize the data structure to improve cache occupancy.

Prerequisite Courses: COE 420

Required Skills: Experience with C/C++ with OpenMP on MIC

Outcome: An N-Body parallel program with shorter execution time due to improved data structure and the use of tiling of the massive data.

Project Requirements:

  • Show impact on program time of using different scenarios for tiling the massive data.
  • Show impact on program time of using improved data structure.
  • Demonstrate significant drop in parallel program time as a result of using some tiling approaches and enhanced data structure.
Advisor: Dr. Mayez Al-Mouhamed No. of students: 3

Satellite Monitoring of Wireless Sensor Networks (WSNs) for Gas detection

Problem Statement: There is huge demand for efficient gas detection in industrial environments for safety and regulatory purposes. Wireless technology is quickly replacing expensive and inflexible wired detection systems. How can we use wireless sensor networks (WSN) for gas detection systems using RF and satellite technology? Find an appropriate gas detection sensor which can work with low-power, low-cost sensor nodes? How to interactively present the information from gas sensors to easily analyze the data?

Prerequisite Courses: COE 241 or COE 344 or COE 482

Required Skills: Hardware, programming, web development

Outcome: A WSN-based gas detection system prototype. An interactive website.

Project Requirements:

  • Interface of a gas sensor with sensor nodes (hardware/software).
  • Interface with a WSN and programming of the satellite terminal.
  • XML gateway interface to receive and process messages.
  • An interactive web application.
Advisor: Dr. Tarek Sheltami No. of students: 4