James R. Green  PhD (Queen's), P.Eng., SMIEEE

Associate Professor

James Green - 4th year projects

4th Year Projects

Please see the SYSC4907/4917 course website for a list of currently available projects. For an idea of the types of projects I am willing to supervise, please see the list of past projects below.

If you are interested in working with me during your 4th year project, please contact me at jrgreen@sce.carleton.ca. Please include the following information for all prospective group members:

  1. A resume/C.V. showing both academic and industry experience
  2. An academic audit (electronic)
  3. A brief statement of interest detailing which project(s) you are interested in working on, and what you hope to gain from this experience (e.g. software vs. hardware experience, research experience, collaboration with industry or other departments, honing a particular skill, etc.)

Current Projects

None yet...

Past Projects

Electronic swimming coach for blind athletes (2009-2010)
Students:  Jaclyn Baldwin, David Galarneau, Adam Jones, and Alexa Loiskandl
Desc: Swimmers who are visually impaired require a coach to manually indicate when they are approaching the ends of the pool and also to keep them centered in the lane. An electronic system is desired to perform these duties. This challenging project will begin by identifying a suitable sensor technology and will culminate with a working waterproof prototype. This project extended the project by the same name below in the areas of waterproofing, human trials, and porting the software to Python. This project was completed in collaboration with Dr. Andrew Marble (SCE).

Application of Sensor Networks in a Smart Apartment (2009-2010)
Students:  Furkan Alaca and Omid Salehi-Abari
Desc: This project explored the use of a wireless sensor network for promoting independent living by tracking various indicators of general human health. This research was completed under teh co-supervision of Prof Rafik Goubran. This project resulted in a paper presented at CMBEC.

Robotic Guide Dog (2009-2010)
Students:  Ronnie Farrel, Manish Walia, and Taha Lukmanji
Desc: This project represents the first year in a multi-year initiative to create a robotic guide dog (RGD). The robotic platform, microcontroller board, and speech recognition subsystems were developed. This project resulted in a paper presented at CMBEC.

Using the Cell BE processor for mass spectrometry (2009-2010)
Students:  Faizan Sultan, Suchita Kannangara, and Kasun Wijenayake
Desc: A continuation of the project of the same name below, this project aimed to preprocess mass spectrometry data and implement a spectral cross-correlation algorithm on the Cell BE with the ultimate goal of completing computational mass spectrometry on the Cell BE processor.

Using the Cell BE processor for nonlinear system identification (2008-2009)
Students:  Ravishankar Kanagasundaram and Rahul Rohra
Desc: Although originally designed for multimedia-intensive gaming on the Sony PS3, the IBM Cell BE processor is also very well suited for computationally intensive scientific computing applications. For example, FFT has been shown to run up to 100x faster on a Cell than on an Opteron with equivalent clock frequency. Since the Cell BE is a heterogeneous multi-core chip, an entirely new programming paradigm is required to effectively utilize its copious computing resources. In this project, a method of nonlinear system identification will be ported to the Cell to achieve massive speedup. This system has applications in biomedical engineering and also in bioinformatics.

Using the Cell BE processor for mass spectrometry (2008-2009)
Students:  Hanan Mahmoud and Robert Peace
Desc: Given a mixture of unknown proteins such as a blood sample, mass spectrometry (MS) can be used to identify the mass/charge ratio of each protein fragment following protein digestion by a suitable restriction enzyme. Much like reassembling a shredded document, the goal of this project is to develop a probabilistic system to identify the original constituent proteins from the individual protein fragments. The Cell BE will be used to dramatically accelerate the protein identification stage such that real-time control of the mass spectrometry device is achieved. In this way, hypothesis-driven data collection may be achieved.

Electronic swimming coach for blind athletes (2008-2009)
Students:  Ramzi Marjaba, Catalin Patulea, and Trevor Gelowski
Desc: Swimmers who are visually impaired require a coach to manually indicate when they are approaching the ends of the pool and also to keep them centered in the lane. An electronic system is desired to perform these duties. This challenging project will begin by identifying a suitable sensor technology and will culminate with a working waterproof prototype. It is hoped to work with a real client on this project. This project will be completed in collaboration with Dr. Andrew Marble (SCE).

Smart Rollator (2008-2009)
Students: Rizwan Haider, Andy Jung, and Christopher Arksey
Desc: In collaboration with Dr. Adrian Chan (Carleton, SCE), this project will continue to develop the usage monitoring subsystem of the Smart Rollator.
Red-Light/Green-Light Playing Robot (2007-2008)
Students: Paola Osorio and Peter Wiebe
Desc: Create a simple robot that detects approaching children and sprays them with 'silly string'. A game of red-light/green-light could be implemented by having the robot “turn its back” to the children, and periodically check to see if they are moving. A vision system, possibly based on an infrared scanner, will be required. Note that a pest deterrent system based on the same principles is commercially available (without robotics).

Behaviour modification via machine vision (2007-2008)
Students: Faezeh Rafsanjani-Sadeghi and WeiZhong Li
Desc: In collaboration with Dr. Andy Adler (Carleton, SCE), this project seeks to develop a prototype system that will monitor a child's activities using a web camera and disable the television when inappropriate behaviour is observed. For example, if a child places his fingers in his mouth, then the TV will be disabled for some period of time. It is hoped that use of such a system will be an effective means of achieving behaviour modification.

Using the Cell BE processor for nonlinear system identification (2007-2008)
Students:  Payam Belkameh and Mohyeldin Hamdy
Desc: Although originally designed for multimedia-intensive gaming on the Sony PS3, the IBM Cell BE processor is also very well suited for computationally intensive scientific computing applications. For example, FFT has been shown to run up to 100x faster on a Cell than on an Opteron with equivalent clock frequency. Since the Cell BE is a heterogeneous multi-core chip, an entirely new programming paradigm is required to effectively utilize its copious computing resources. In this project, a method of nonlinear system identification will be ported to the Cell to achieve massive speedup. This system has applications in biomedical engineering and also in bioinformatics. Note that this project led to a poster presentation at the IBM Cell BE Programming Workshop held at Carleton University May 15-16 2008

Vehicle Anit-Doze System (2007-2008)
Students:  Eric Au and Kevin Charland
Desc: By tracking user gaze using infrared, this system will detect when a driver is showing drowsy behaviour and will warn the user. Via an LED prompt, the user will first be asked to check their mirror. If the user fails to respond to this prompt, a vibration module in the driver seat will be activated and an audible warning system will be activated.

Eye Interact (2007-2008)
Students: Ryan Steeves and Michael Colussi
Desc: In collaboration with Dr. Adrian Chan (Carleton, SCE), eye-Interact is a low-cost, eye movement controlled communication system for persons with high-level physical disabilities. As extension of a previous team's efforts, this project aims to track user gaze using an infrared camera and a system of infrared LEDs via the cross-ratios method.

Intelligent Systems for Bioinformatics: Protein Analysis Tool (2007-2008)
Students: Kagiso Mguni
Desc: In collaboration with Dr. Susan Aitken (Carleton, BIO), this project will extend previous work by Greg Dmochowski to build a protein analysis tool. A Java program will be extended to automate an analysis pipeline that aims to identify key residues that confer function. These residues will be identified through comparative analysis between two or more families of proteins with representative sequences from multiple species.

Smart Rollator: Obstacle Avoidance and Guidance System (2007-2008)
Students: Mohammed Aboul-Magd, Faysal Hasan and Alex Sintu
Desc: In collaboration with Dr. Adrian Chan (Carleton, SCE), this project will provide the Smart Rollator with a means to advise the user of upcoming obstacles and will guide the user to selected locations such as their room or the cafeteria. Sensors include infrared and sonar. See the Wiki site for details.

Smart Rollator: Usage Monitoring (2007-2008)
Students: Davide Agnello and Brian Earl
Desc: In collaboration with Dr. Adrian Chan (Carleton, SCE), this project will monitor usage of the Smart Rollator including distance travelled, total usage time, and average speed. Sensors will be interfaced to an embedded system and data will be logged in nov-volatile memory. Updates will be sent periodically via a BlueTooth interface to a central data repository. Different roles (e.g. patient, doctor, family) will see different views of patient data. See the Wiki site for details. Note that this project was awarded the NSERC best project award and a prototype is now being tested by physiotherapy students and real patients.

Assistive device for children with Down syndrome (2006-2007)
Students: Jonathan LaRocque and Kyle Mulligan
Desc: Children born with Down syndrome often have low muscle tone which may affect the muscles of the mouth. This sometimes leads to the appearance of a protruding tongue. The goal of this project was to develop an interactive computer system that will lead to increased proprioception (positional awareness), strength, and dexterity of the tongue. The system will consist of two principle components: 1) a safe non-invasive tongue tracking system, possibly based on a webcam; and 2) an interactive program (game) that will engage the child and lead the child through a series of exercises. This project resulted in a fully functional tongue-tracker prototype and a conference paper presented at CMBEC30 in Toronto.

Smart Walker: Force and Gait Analysis (2006-2007)
Students: James Makienko and George Shenouda
Desc: In collaboration with Dr. Adrian Chan (Carleton, SCE), a rollator (wheeled walker) was outfitted with accelerometers, force sensors, and wheel encoders in order to analyze a user's gait during use. It is hoped that such a system will help the elderly to live independently.

Smart Walker: Robust Heart Rate Detection via Data Fusion (2006-2007)
Students: Lisa Boyachok, Yasmin Khezri, and Jed Vito
Desc: In collaboration with Dr. Adrian Chan (Carleton, SCE), a data fusion algorithm and GUI was developed to intelligently combine ECG and PPG data to achieve a robust measure of heart rate. Ultimatley, this system will be integrated into the Smart Walker for remote monitoring of elderly patients' health.

Eye Interact: A Low-Cost Eye Movement Controlled Communication System (2006-2007)
Students: Amir Sadeghian and Ryan (Chol-ho) Yim
Desc: In collaboration with Dr. Adrian Chan (Carleton, SCE), students developed a low-cost, eye movement controlled communication system for persons with high-level physical disabilities. This system consists of a computer equipped with a webcam to achieve gaze tracking in order to activate menu items. This project was awarded the NSERC best project award, resulted in a fully functional gaze-tracking system and a conference paper presented at CMBEC30 in Toronto.

Towards Protein Structure: The prediction of protein domain boundaries (2005-2006)
Student: Sankua Chao
Desc: In collaboration with Dr. Michel Dumontier (Carleton, Biology), a computer system was developed for the prediction of protein domain boundaries based solely on protein sequence data. This study not only reproduced a recently published method, but extended the system by making use of artificial neural networks to create more complex and precise predictors.