Innovative Projects Realized

Thermoelectric Applications in Waste Heat Recovery

The student will conduct a detailed literature review on the subject, derive a theory for power generation, and construct a simple test apparatus to evaluate the theory. 

Design and Construction of Gust Simulator for Wind-Turbine Testing

Vibration Assisted Nano Mechanical Machining through Atomic Force Microscope

The student will perform tribological tests in nano scale through force spectroscopy, scratch and nano-indentation tests to identify flow stress and elasto-plastic tribological properties, such as hardness, the friction coefficient, elastic recovery, and plowing effects. The second part of the work is an implementation of nano stage which we have developed in our laboratory using piezo-electric actuator. The student will investigate the dynamic characteristics of the piezoelectric actuator stage.  A real-time operating and control system (DSpace 1103PX4) will be used to actuate the elliptical shapes through a Matlab/SimulinkTM environment. 

Partial differential equations for seismic imaging.

Parallel Software Acceleration of Meta-Heuristics for Data Mining

Over the last ten years, the needs of industry have made data mining one of the most important facets of Information Technology (IT). In simple terms, data mining is the automatic process of extracting interrelationships and patterns of interest from data. Today, companies around the world rely on data mining not only to discover knowledge from historical information, but to predict future  trends and behaviors, allowing businesses to make more informed decisions.

The goal of this project is two-fold: (1) To construct scalable meta-heuristics by exploiting the parallelism that is available in today’s commodity hardware; and (2) use these scalable, parallel meta-heuristics as part of an accelerator for data mining. With regards to (1) we propose to consider only populationbased metaheuristics, as these lend themselves most naturally to parallel  implementations. The metaheuristics that will be considered include genetic algorithms, greedy randomized adaptive search
procedures, estimation of distribution algorithms, particle swarm optimization, and ant colony algorithms. When seeking to parallelize these algorithms we plan to only exploit commodity hardware in the form of multi-core and many-core architectures. The decision to limit ourselves to commodity hardware makes our algorithms immediately accessible to anyone with a desktop PC with a Graphics Processing Unit (GPU). Currently, the latter contains an enormous amount of potential computing power with (simple) GPU cores outnumbering (complex) CPU cores by two orders-of-magnitude. By implementing the meta-heuristics directly on the GPU or a combination of CPU-GPU, we hope to achieve a speedup of at least one order-of-magnitude over software. Once complete, the metaheuristics from (1) will be used to develop a scalable, parallel accelerator for data mining. Our shortterm (12-week) goal is to use these parallel metaheuristics to construct scalable classifiers.

The proposed project research program consists of theoretical analysis supplemented with experimental prototyping. Thus, the student will not only gain knowledge in formal thinking, but also practical skills in developing complex parallel software systems.

Structure and Function of a Ribonucleoprotein Particle

The Kothe research group studies the structure and function of small ribonucleoproteins which are involved in ribosome biogenesis, in particular in modification of ribosomal RNA, with the aim of better understanding how the cell assembles large RNAprotein complexes such as the ribosome.

My basic research investigates the mechanism of the early stages in ribosome biogenesis on a molecular level. For this aim, I am applying a powerful, multidisciplinary combination of traditional yeast genetics and molecular biology with state-of-the-art biochemistry and biophysics using bacterial, archaeal and eukaryotic model systems.

The Globalink student will assist in all stages of the project allowing him to get detailed experience in many areas of  biochemistry and molecular biology. One of our main questions addresses the function of certain structural elements of the ribonucleoprotein particle for example for substrate recognition or complex formation. To answer this question, we alter specific protein or RNA regions using site-directed mutagenesis methods. During his/her internship, the Globalink student will use these techniques to generate mutations in the DNAs encoding the protein components of the ribonucleoproteins complex. In order to reconstitute the mutant ribonucleoprotein particle, the mutant protein needs to be recombinantly overexpressed in E. coli and subsequently be purified by affinity and other types of chromatography.
Also the other proteins of the ribonucleoproteins complex will be purified by similar methods. The success of the protein purifications is followed by SDS-PAGE. Next, the RNA component of the ribonucleoproteins complex is generated by in vitro transcription followed by purification using anion exchange chromatography which is monitored by RNA-Urea-PAGE. Lastly, the
student will analyze formation of the ribonucleoprotein complex by size-exclusion chromatography comparing both the wild-type and the mutant components. Together with my Ph.D. student, the Globalink student will also have the opportunity to assess effects of the introduced mutation on the ability of the ribonucleoproteins particle to modify RNA. In summary, this project aims at furthering our understanding of ribonucleoprotein structure and function in ribosome biogenesis. This knowledge will be important to target this critical cellular pathway in different diseases.

Strengthening Prestressed/Precast Hollow Core Slabs with Near-Surface Mounted FRP Reinforcement

Sensor design and historical data analysis for condition monitoring of overhead conductors

Experimental investigation of anaerobic digestion conducted under controlled conditions: data collection and analysis

Anaerobic digestion is an environmentally-friendly process that breaks down biomass into simpler, more stable components, while simultaneously capturing energy in the form of a methane-rich biogas.  Digesters are usually used to treat wastes, such as at municipal waste facilities, or from livestock farms.  However, anaerobic digesters are not very reliable, suffering from occasional “digester upsets”.  Consequently, anaerobic digestion is less attractive to industry as compared to other less environmental alternatives.  The reason behind this unreliability lies in the complexities of anaerobic digestion – the digestion process involves hundreds of different species of micro-organisms, each with their own set of interrelated behaviours.  The purpose of our project is to develop a more reliable 3-D computer model on which to base future digester designs and control systems.  The result of this is a general improvement in the reliability of anaerobic digesters, leading to an increased installation rate in industry.

The student is expected to perform data gathering activities, the nature of which will vary depending on the controlled experiments to be performed and the facility. Flow visualization activities involves becoming familiar with the operation of a fluid column system, the Particle Image Velocimeter and the Acoustic Doppler Velocimeter , conducting experimental trials, and performing initial data analysis, such as software-assisted adaptive correlation, as well as using MATLAB to produce results. Activities at the pilot-scale digester facility include daily recording of sensor readings, and address control issues indentified by the controller.  Data reduction can be expected using Excel and made ready to validate against a 3-D anaerobic digester model.

Improving Signature-Based Intrusion Detection

The prevalent use of signature-based approach in modern intrusion detection systems (IDS) emphasizes the importance of two issues associated with the performance of the approach: the rigid requirements for signature processing and the quality of signature set. The focus of this research project is on improving performance of signature-based intrusion detection.

The signature-based intrusion detection is one of the most commonly used techniques implemented in modern intrusion detection systems (IDS). This method is based on matching incoming events against a set of rules, i.e., attack signatures, to identify known intrusions. One of the major advantages of signature-based approach and the primary reason for its widespread acceptance is the predictability of its behaviour and accuracy of detection. On the other hand, there are two issues associated with this approach in practice: the performance of the matching component and the quality of the signature set.

The performance has long become a critical operation in the signature-based intrusion detection systems. The rigid performance requirements are dictated not only by increasing network speeds but also by increasing complexity and quantity of intrusion detection signatures. Constant discovery of the software vulnerabilities and novel threats demands timely addition of new attack signatures. This results in a complex, overlapping and often redundant set of attack signatures.

The aim of this research project is therefore two-fold: to develop an efficient signature matching approach and tools for analysis of the signature set quality.

Experimental Computer-Aided Design Flow

Field-Programmable Gate Arrays (FPGAs) have become the implementation medium of choice for many digital circuits in areas as diverse as telecommunications, bioinformatics, visualization systems, and digital signal processing. One of the few areas where FPGAs have not yet become ubiquitous is in mobile applications.This application area is huge; together with cloud computing, we expect that mobile devices will become the computing platform of the future. These devices will represent much more than email and web platforms; they will provide personal and immediate computing applications that will aid, heal, inform and empower humanity in the next few decades. The primary reasons that Field-Programmable Gate Arrays have not been employed in mobile devices is their energy dissipation and cost. In this project, we will investigate (a) new energyefficient FPGA architectures, (b) a method for mapping applications to FPGAs with energy as a first-class concern, and (c) an understanding and demonstration of what is possible in a mobile device enhanced with our programmable platform. We have partnered with Altera, a major FPGA vendor with a significant R&D presence in Toronto, who would be able to immediately employ our techniques to produce better FPGAs.
We will explore and develop energy-aware CAD algorithms and software tools that can map circuits to the architectures proposed in (a). We will engage in an extensive collaborative effort, building on [1] whose purpose is to enable FPGA architecture and CAD research based on applications designed at the Verilog HDL level. This CAD flow consists of HDL elaboration, logic optimization, technology mapping, clustering, placement, and routing. Our prior research has identified energy-efficient methods for several of these steps [SW15;2]; however, this did not consider the novel features proposed in Task 1. In addition, we have not considered energyefficient elaboration, which is essential to make effective use of these energy-savings features. The energy-savings features of an FPGA must be designed to be general-purpose and suit as many applications as possible. This is a fundamental difference between reducing power in a fixed-function chip compared to an FPGA. For a fixed-function chip, the designer can examine the application, and create voltage islands or other power-saving structures that are optimized specifically for that application. When mapping the application to an FPGA, however, the
designer must use the collection of power-savings features available on the FPGA in a judicious manner. In this project, we will seek methods to automate the translation of the application’s power requirements to the implementation of those requirements using the available FPGA poweroptimization features. The methods will be encapsulated in a Power Elaboration algorithm and tool.

Isolation of hemicelluloses from the kraft-based dissolving pulp production process

The three main components in wood chips are cellulose, hemicelluloses, and lignin. Traditionally, cellulose is the main constitute of pulp, which is the product of pulp mills and is used as the main raw material for paper production in paper mills. The lignin, however, is mainly burned to supply the heat requirement of the pulp mills.

However, the hemicelluloses are conventionally burned along with lignin in the pulping spent liquor recovery system.  Since the heat values of hemicelluloses are much lower than that of the dissolved lignin in the spent liquor, it is widely accepted that burning is not the economical utilization of hemicelluloses. Researchers are seeking for alternative methods to improve the utilization of hemicelluloses, which in turn increases the revenue of the pulp mills.

The student internship will be directly involved in the research program.  The candidate will learn state of the art analytical and processing techniques, and will be trained to understand and carry out innovative research program. He/She will also develop teamwork, leadership presentation and communication skills.